docs: 添加 ablation 消融实验 + README 版本概览

README.md

@@ -0,0 +1,32 @@
+# Beast Trader 策略仓库
+
+ETH/USDT 永续合约量化交易策略版本管理，基于 freqtrade + Binance。
+
+## 当前部署
+
+**v2.2d** — 三层趋势共振（D1+4H+1H），震荡市自动休眠，dry-run 运行中
+
+## 版本演进
+
+| 系列 | 版本范围 | 方向 | 状态 |
+|------|---------|------|------|
+| v0.x | v0.1 ~ v0.3 | 价格行为探索 | 已弃用 |
+| v1.x | v1.0 ~ v1.9 | 结构流策略迭代 | 已弃用 |
+| v2.x | v2.0 ~ v2.2d | 趋势跟踪（当前主线） | **v2.2d 运行中** |
+| v3.x | v3.0 ~ v3.2 | 震荡波段（Swing） | 已验证/备用 |
+| v4.x | v4.0 ~ v4.2 | 极简震荡 | 实验 |
+| Scalp | v1.8, v2.0 | 剥头皮 | 已弃用 |
+
+## 关键教训
+
+- v1.1~v1.8 Scalp：反向S/R交易 = 逆势接飞刀（0%胜率）
+- v2.3：参数调优不是方向（创建后10分钟删除）
+- v2.2b：当前最优回测基线（+4673%/+17%最大回撤）
+- v2.2d：D1趋势总闸门 — 震荡市不下单是保护机制不是bug
+
+## 铁律
+
+1. 只增不删 — 所有历史版本保留
+2. 版本归档 — 每个版本独立 commit
+3. 回测标准化 — 复用成功配置
+4. 主任不越俎代庖 — 方案设计归主任，代码编写归交易部

ablation/ablation_1.py

@@ -0,0 +1,454 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl1(IStrategy):
+    """
+    Ablation Variant 1: 移除条件 1
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["in_demand_4h"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist <= max_dist)
+            & (long_stop_dist > 0.003)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["in_supply_4h"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist <= max_dist)
+            & (short_stop_dist > 0.003)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_2.py

@@ -0,0 +1,454 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl2(IStrategy):
+    """
+    Ablation Variant 2: 移除条件 2
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist <= max_dist)
+            & (long_stop_dist > 0.003)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist <= max_dist)
+            & (short_stop_dist > 0.003)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_3.py

@@ -0,0 +1,454 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl3(IStrategy):
+    """
+    Ablation Variant 3: 移除条件 3
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["in_demand_4h"]
+            & (long_stop_dist <= max_dist)
+            & (long_stop_dist > 0.003)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["in_supply_4h"]
+            & (short_stop_dist <= max_dist)
+            & (short_stop_dist > 0.003)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_4.py

@@ -0,0 +1,454 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl4(IStrategy):
+    """
+    Ablation Variant 4: 移除条件 4
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["in_demand_4h"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist > 0.003)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["in_supply_4h"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist > 0.003)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_5.py

@@ -0,0 +1,454 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl5(IStrategy):
+    """
+    Ablation Variant 5: 移除条件 5
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["in_demand_4h"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist <= max_dist)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["in_supply_4h"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist <= max_dist)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_6.py

@@ -0,0 +1,454 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl6(IStrategy):
+    """
+    Ablation Variant 6: 移除条件 6
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["in_demand_4h"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist <= max_dist)
+            & (long_stop_dist > 0.003)
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["in_supply_4h"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist <= max_dist)
+            & (short_stop_dist > 0.003)
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_7.py

@@ -0,0 +1,454 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl7(IStrategy):
+    """
+    Ablation Variant 7: 移除条件 7
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["in_demand_4h"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist <= max_dist)
+            & (long_stop_dist > 0.003)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["in_supply_4h"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist <= max_dist)
+            & (short_stop_dist > 0.003)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_8.py

@@ -0,0 +1,456 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Abl8(IStrategy):
+    """
+    Ablation Variant 8: 移除条件 8
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["in_demand_4h"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist <= max_dist)
+            & (long_stop_dist > 0.003)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = True  # cooldown removed
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["in_supply_4h"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist <= max_dist)
+            & (short_stop_dist > 0.003)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = True  # cooldown removed
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/ablation_all_removed.py

@@ -0,0 +1,442 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21_Ablall(IStrategy):
+    """
+    Ablation Variant all: 移除条件 1,2,3,4,5,6,7,8
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+        )
+
+        long_recent = True  # cooldown removed
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+        )
+
+        short_recent = True  # cooldown removed
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

ablation/v2_1_baseline.py

@@ -0,0 +1,456 @@
+"""
+Structure Flow Strategy v2.1
+=======================
+变更记录：
+  v1.6 (2026-06-07): 最优基线 — +3659.63%, 190笔, 69.3% trailing胜率
+  v2.0 (2026-06-08): B1 入场延迟确认 — 方向正确但降频严重
+  v2.1 (2026-06-08): ===== D1: 趋势强度过滤 =====
+                      在4H级别评估趋势强度：最近2个Swing Point的间距变化。
+                      如果趋势在扩张（HH/HL间距增大），允许入场；
+                      如果趋势在收缩（HH/HL间距缩小或震荡），过滤信号。
+                      目的：只在趋势明确时交易，避免震荡市反复止损。
+"""
+
+from datetime import datetime
+import numpy as np
+import pandas as pd
+from pandas import DataFrame
+from freqtrade.strategy import IStrategy, IntParameter, informative
+from freqtrade.persistence import Trade
+
+
+class StructureFlowStrategyV21(IStrategy):
+    """
+    Structure Flow Strategy v2.1 — D1: 趋势强度过滤
+
+    v2.1改动（相对于v1.6）：
+      在4H级别计算趋势强度：最近2个Swing High间距 + Swing Low间距的变化。
+      只有趋势在扩张（或至少不收缩）时才允许入场。
+    """
+
+    can_short = True
+    stoploss = -0.15
+    use_custom_stoploss = True
+    minimal_roi = {"0": 100}
+    max_open_trades = 1
+    timeframe = "1h"
+
+    # =====================
+    # 可优化参数
+    # =====================
+
+    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
+    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
+    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
+    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
+    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
+    # v2.1 新增：趋势强度最小扩张比例（x/100 = 0%~50%）
+    # 0 = 只要不收缩就行；越大要求趋势扩张越强
+    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")  # -20=允许SP轻微收缩, 最佳值
+
+    # =====================
+    # 工具：Swing Point 检测
+    # =====================
+
+    @staticmethod
+    def _detect_swing_points(
+        high: pd.Series,
+        low: pd.Series,
+        window: int = 5,
+    ) -> tuple[pd.Series, pd.Series]:
+        n = len(high)
+        sh = pd.Series(np.nan, index=high.index, dtype=float)
+        sl = pd.Series(np.nan, index=low.index, dtype=float)
+
+        for i in range(window, n - window):
+            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
+                sh.iloc[i] = high.iloc[i]
+            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
+                sl.iloc[i] = low.iloc[i]
+
+        return sh, sl
+
+    # =====================
+    # 工具：结构分析
+    # =====================
+
+    def _build_structure(
+        self,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        swing_high: pd.Series,
+        swing_low: pd.Series,
+    ) -> DataFrame:
+        n = len(high)
+
+        trend_up_arr = np.full(n, False)
+        trend_down_arr = np.full(n, False)
+        nearest_support = np.full(n, np.nan)
+        nearest_resistance = np.full(n, np.nan)
+        in_demand_zone = np.full(n, False)
+        in_supply_zone = np.full(n, False)
+
+        sh_prices = []
+        sl_prices = []
+
+        for i in range(n):
+            if pd.notna(swing_high.iloc[i]):
+                sh_prices.append(swing_high.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+
+            if pd.notna(swing_low.iloc[i]):
+                sl_prices.append(swing_low.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
+                    trend_up_arr[i] = True
+                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
+                    trend_down_arr[i] = True
+                elif i > 0:
+                    trend_up_arr[i] = trend_up_arr[i - 1]
+                    trend_down_arr[i] = trend_down_arr[i - 1]
+            elif i > 0:
+                trend_up_arr[i] = trend_up_arr[i - 1]
+                trend_down_arr[i] = trend_down_arr[i - 1]
+
+            if sl_prices:
+                nearest_support[i] = sl_prices[-1]
+            if sh_prices:
+                nearest_resistance[i] = sh_prices[-1]
+
+            c = close.iloc[i]
+            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
+                zone_range = nearest_resistance[i] - nearest_support[i]
+                if zone_range > 0:
+                    pos_pct = (c - nearest_support[i]) / zone_range
+                    in_demand_zone[i] = pos_pct < 0.35
+                    in_supply_zone[i] = pos_pct > 0.65
+
+        return DataFrame({
+            "trend_up": trend_up_arr,
+            "trend_down": trend_down_arr,
+            "support": nearest_support,
+            "resistance": nearest_resistance,
+            "in_demand": in_demand_zone,
+            "in_supply": in_supply_zone,
+        }, index=high.index)
+
+    # =====================
+    # 工具：K线形态检测
+    # =====================
+
+    @staticmethod
+    def _detect_candle_patterns(
+        open_: pd.Series,
+        high: pd.Series,
+        low: pd.Series,
+        close: pd.Series,
+        pin_bar_wick_ratio: float = 0.6,
+    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
+        body = (close - open_).abs()
+        total_range = (high - low).replace(0, 0.0001)
+
+        upper_wick = high - close.where(close > open_, open_)
+        lower_wick = open_.where(close > open_, close) - low
+        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
+
+        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
+        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
+
+        prev_open = open_.shift(1)
+        prev_close = close.shift(1)
+        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
+        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
+
+        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
+
+    # ================================================================
+    #  信息时间框架 — D1 宏观结构
+    # ================================================================
+
+    @informative("1d")
+    def populate_indicators_1d(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_d1.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        return dataframe
+
+    # ================================================================
+    #  信息时间框架 — 4H 中期结构
+    # ================================================================
+
+    @informative("4h")
+    def populate_indicators_4h(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        sh, sl = self._detect_swing_points(
+            dataframe["high"], dataframe["low"],
+            self.swing_lookback_h4.value,
+        )
+        structure = self._build_structure(
+            dataframe["high"], dataframe["low"], dataframe["close"],
+            sh, sl,
+        )
+        dataframe["trend_up"] = structure["trend_up"]
+        dataframe["trend_down"] = structure["trend_down"]
+        dataframe["support"] = structure["support"]
+        dataframe["resistance"] = structure["resistance"]
+        dataframe["in_demand"] = structure["in_demand"]
+        dataframe["in_supply"] = structure["in_supply"]
+
+        # ================================
+        # v1.6 活支撑/阻力检查（保留）
+        # ================================
+        touched_support = (
+            (dataframe["low"] <= dataframe["support"] * 1.005) &
+            (dataframe["low"] >= dataframe["support"] * 0.995)
+        )
+        held_support = dataframe["close"] > dataframe["support"]
+        support_tested_and_held = touched_support & held_support
+        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        touched_resistance = (
+            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
+            (dataframe["high"] <= dataframe["resistance"] * 1.005)
+        )
+        held_resistance = dataframe["close"] < dataframe["resistance"]
+        resistance_tested_and_held = touched_resistance & held_resistance
+        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
+
+        # ================================
+        # v2.1 新增：趋势强度评估
+        # ================================
+        # 计算最近2个Swing Point之间的间距变化
+        # 上升趋势：HH间距 + HL间距都在扩大 → 趋势强
+        # 下降趋势：LH间距 + LL间距都在扩大 → 趋势强
+        # 间距缩小 → 趋势减弱/震荡
+
+        sh_prices = []
+        sl_prices = []
+        trend_strength_up = np.full(len(dataframe), np.nan)
+        trend_strength_down = np.full(len(dataframe), np.nan)
+
+        for i in range(len(dataframe)):
+            if pd.notna(sh.iloc[i]):
+                sh_prices.append(sh.iloc[i])
+                if len(sh_prices) > 4:
+                    sh_prices.pop(0)
+            if pd.notna(sl.iloc[i]):
+                sl_prices.append(sl.iloc[i])
+                if len(sl_prices) > 4:
+                    sl_prices.pop(0)
+
+            # 上升趋势强度：HH[-1] vs HH[-2], HL[-1] vs HL[-2]
+            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
+                # HH间距：最近两个Swing High的差值百分比
+                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
+                # HL间距：最近两个Swing Low的差值百分比
+                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
+                # 上升趋势强度 = HH间距 + HL间距（都正=扩张，一正一负=不确定，都负=收缩）
+                trend_strength_up[i] = hh_dist + hl_dist
+
+                # 下降趋势强度（取反：间距缩小是负值）
+                trend_strength_down[i] = -(hh_dist + hl_dist)
+
+        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
+        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
+
+        # 趋势强度是否足够（扩张中）
+        min_strength = self.trend_strength_min.value / 100.0  # 0~0.30
+        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
+        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+
+        return dataframe
+
+    # ================================================================
+    #  主时间框架 — 1H 指标
+    # ================================================================
+
+    def populate_indicators(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """1H 级别：K线形态（零指标）。"""
+        bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
+            self._detect_candle_patterns(
+                dataframe["open"],
+                dataframe["high"],
+                dataframe["low"],
+                dataframe["close"],
+                self.pin_bar_wick_ratio.value / 100.0,
+            )
+        )
+        dataframe["bullish_pinbar"] = bullish_pin
+        dataframe["bearish_pinbar"] = bearish_pin
+        dataframe["bullish_engulfing"] = bullish_engulf
+        dataframe["bearish_engulfing"] = bearish_engulf
+        dataframe["bullish_signal"] = bullish_pin | bullish_engulf
+        dataframe["bearish_signal"] = bearish_pin | bearish_engulf
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        return dataframe
+
+    # =====================
+    # 入场信号
+    # =====================
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        入场逻辑（1H 时间框架）。
+
+        v2.1 核心改动：D1 — 趋势强度过滤
+          做多额外条件：4H上升趋势在扩张（strong_uptrend_4h）
+          做空额外条件：4H下降趋势在扩张（strong_downtrend_4h）
+        """
+        max_dist = self.max_stop_dist.value / 100.0
+        cooldown = self.cooldown_bars.value
+
+        # NaN 安全处理
+        bool_cols = [
+            "trend_up_1d", "trend_down_1d",
+            "trend_up_4h", "trend_down_4h",
+            "in_demand_4h", "in_supply_4h",
+            "support_alive_4h", "resistance_alive_4h",
+            "strong_uptrend_4h", "strong_downtrend_4h",
+            "bullish_signal", "bearish_signal",
+        ]
+        for col in bool_cols:
+            if col in dataframe.columns:
+                dataframe[col] = dataframe[col].fillna(False)
+
+        # ── 做多 ──
+        long_stop_dist = (dataframe["open"] - dataframe["support_4h"]) / dataframe["open"]
+
+        long_base = (
+            dataframe["trend_up_1d"]
+            & dataframe["in_demand_4h"]
+            & dataframe["bullish_signal"]
+            & (long_stop_dist <= max_dist)
+            & (long_stop_dist > 0.003)
+            & dataframe["support_alive_4h"]
+            # v2.1: 趋势强度 — 4H上升趋势必须在扩张
+            & dataframe["strong_uptrend_4h"]
+        )
+
+        long_recent = long_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[long_base & long_recent, "enter_long"] = 1
+
+        # ── 做空 ──
+        short_stop_dist = (dataframe["resistance_4h"] - dataframe["open"]) / dataframe["open"]
+
+        short_base = (
+            dataframe["trend_down_1d"]
+            & dataframe["in_supply_4h"]
+            & dataframe["bearish_signal"]
+            & (short_stop_dist <= max_dist)
+            & (short_stop_dist > 0.003)
+            & dataframe["resistance_alive_4h"]
+            # v2.1: 趋势强度 — 4H下降趋势必须在扩张
+            & dataframe["strong_downtrend_4h"]
+        )
+
+        short_recent = short_base.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[short_base & short_recent, "enter_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 出场信号
+    # =====================
+
+    def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """出场逻辑 — 由结构反转触发。"""
+        exit_long = ~dataframe["trend_up_1d"].fillna(True)
+        dataframe.loc[exit_long, "exit_long"] = 1
+
+        exit_short = dataframe["trend_up_1d"].fillna(False)
+        dataframe.loc[exit_short, "exit_short"] = 1
+
+        return dataframe
+
+    # =====================
+    # 动态止损 — 纯价格结构（基于Swing Point）
+    # =====================
+
+    def custom_stoploss(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        after_fill: bool,
+        **kwargs,
+    ) -> float:
+        """
+        止损逻辑：完全基于价格结构，零指标（与v1.6相同）。
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        last = dataframe.iloc[-1]
+
+        if not trade.is_short:
+            support = last.get("support_4h", np.nan)
+            if pd.isna(support) or support <= 0:
+                return -0.02
+            sl_price = support * 0.999
+            sl_ratio = (sl_price / current_rate) - 1.0
+            return max(sl_ratio, -0.15)
+        else:
+            resistance = last.get("resistance_4h", np.nan)
+            if pd.isna(resistance) or resistance <= 0:
+                return 0.02
+            sl_price = resistance * 1.001
+            sl_ratio = 1.0 - (sl_price / current_rate)
+            return min(sl_ratio, 0.15)
+
+    # =====================
+    # Plot config
+    # =====================
+
+    @staticmethod
+    def plot_config() -> dict:
+        return {
+            "main_plot": {
+                "support_4h": {"color": "green", "type": "line"},
+                "resistance_4h": {"color": "red", "type": "line"},
+            },
+            "subplots": {
+                "signals": {
+                    "bullish_pinbar": {"color": "green", "type": "scatter"},
+                    "bearish_pinbar": {"color": "red", "type": "scatter"},
+                },
+                "filters": {
+                    "support_alive_4h": {"color": "green", "type": "line"},
+                    "resistance_alive_4h": {"color": "red", "type": "line"},
+                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
+                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
+                },
+            },
+        }

strategy.py

@@ -1,398 +1,270 @@
-"""
-Structure Flow Strategy v2.2c — 冷却期修复版
-==============================================
-变更记录：
-  v2.2c (2026-06-11): 1H S/R 替代 4H S/R
-  v2.2c-coolfix (2026-06-11): 修复冷却期无限阻止下单 bug
-"""
+# structure_flow_momentum_scalp.py
+# 顺趋势剥头皮策略 v2.0
+#
+# 核心思路：不再在S/R处做反向交易接飞刀，而是顺趋势方向，等回调后入场。
+#
+# ┌─────────────────────────────────────────────────────────────┐
+# │  15m趋势方向判断（EMA20 vs EMA50）                           │
+# │         ↓                                                    │
+# │  上升趋势 → 只等5m回调到EMA20/支撑附近 → 止跌信号 → 做多     │
+# │  下降趋势 → 只等5m反弹到EMA20/阻力附近 → 止涨信号 → 做空     │
+# │         ↓                                                    │
+# │  止损：ATR×1.0 | 止盈：ATR×1.5 | 时间止损：60分钟           │
+# └─────────────────────────────────────────────────────────────┘
+#
+# v2.0 (2026-06-10): 初始版本，完全重写
 
-from datetime import datetime
-import numpy as np
-import pandas as pd
+from freqtrade.strategy import IStrategy, IntParameter, DecimalParameter, informative
 from pandas import DataFrame
-from freqtrade.strategy import IStrategy, IntParameter, informative
+import pandas as pd
+import numpy as np
+from datetime import datetime
 from freqtrade.persistence import Trade
 
 
-class StructureFlowStrategyV22d(IStrategy):
+class StructureFlowMomentumScalp(IStrategy):
+    """
+    顺趋势剥头皮策略 v2.0
+
+    核心逻辑：
+    - 15m EMA趋势方向过滤，只做顺趋势方向的单
+    - 5m 回调到EMA20或S/R支撑/阻力区域时，等待K线信号确认后入场
+    - 止损 ATR×1.0，止盈 ATR×1.5，时间止损 60 分钟
+    - 不做方向猜测，不吃鱼头鱼尾，只吃回调结束那一小段
+    """
+
+    # ── 时间框架 ──
+    timeframe = "5m"
+
+    # ── 交易参数 ──
     can_short = True
-    stoploss = -0.15
-    use_custom_stoploss = True
-    minimal_roi = {"0": 100}
     max_open_trades = 1
-    timeframe = "1h"
+    stake_amount = "unlimited"
+    use_custom_stoploss = True
+    use_exit_signal = False  # 出场完全由 custom_stoploss + custom_exit 管理
 
-    # =====================
-    # 可优化参数
-    # =====================
+    # ── 合约参数 ──
+    margin_mode = "cross"
+    trading_mode = "futures"
 
-    swing_lookback_d1 = IntParameter(8, 14, default=10, space="buy")
-    swing_lookback_h4 = IntParameter(5, 10, default=8, space="buy")
-    swing_lookback_1h = IntParameter(3, 7, default=5, space="buy")  # 新增：1H swing参数
-    pin_bar_wick_ratio = IntParameter(50, 70, default=60, space="buy")
-    max_stop_dist = IntParameter(20, 50, default=50, space="buy")
-    cooldown_bars = IntParameter(3, 12, default=6, space="buy")
-    trend_strength_min = IntParameter(-50, 20, default=-20, space="buy")
+    # ── 可优化参数 ──
+    # 趋势检测
+    trend_ema_period = IntParameter(10, 30, default=20, space="buy")
+    # 回调确认幅度
+    pullback_deviation = DecimalParameter(0.2, 1.0, default=0.5, decimals=1, space="buy")
+    # 入场冷却期
+    cooldown_bars = IntParameter(2, 8, default=3, space="buy")
+    # K线形态灵敏度
+    pin_bar_wick_ratio = IntParameter(50, 80, default=60, space="buy")
+    # 止损ATR倍数
+    atr_mult_stop = DecimalParameter(0.8, 2.0, default=1.0, decimals=1, space="sell")
+    # 止盈ATR倍数
+    atr_mult_tp = DecimalParameter(1.0, 3.0, default=1.5, decimals=1, space="sell")
 
-    # =====================
-    # 工具：Swing Point 检测
-    # =====================
+    # ── 常数 ──
+    time_stop_minutes = 60  # 最大持仓时间
 
-    @staticmethod
-    def _detect_swing_points(
-        high: pd.Series,
-        low: pd.Series,
-        window: int = 5,
-    ) -> tuple[pd.Series, pd.Series]:
-        n = len(high)
-        sh = pd.Series(np.nan, index=high.index, dtype=float)
-        sl = pd.Series(np.nan, index=low.index, dtype=float)
-
-        for i in range(window, n - window):
-            if high.iloc[i] > high.iloc[i - window : i].max() and high.iloc[i] > high.iloc[i + 1 : i + window + 1].max():
-                sh.iloc[i] = high.iloc[i]
-            if low.iloc[i] < low.iloc[i - window : i].min() and low.iloc[i] < low.iloc[i + 1 : i + window + 1].min():
-                sl.iloc[i] = low.iloc[i]
-
-        return sh, sl
-
-    # =====================
-    # 工具：结构分析
-    # =====================
-
-    def _build_structure(
-        self,
-        high: pd.Series,
-        low: pd.Series,
-        close: pd.Series,
-        swing_high: pd.Series,
-        swing_low: pd.Series,
-    ) -> DataFrame:
-        n = len(high)
-
-        trend_up_arr = np.full(n, False)
-        trend_down_arr = np.full(n, False)
-        nearest_support = np.full(n, np.nan)
-        nearest_resistance = np.full(n, np.nan)
-        in_demand_zone = np.full(n, False)
-        in_supply_zone = np.full(n, False)
-
-        sh_prices = []
-        sl_prices = []
-
-        for i in range(n):
-            if pd.notna(swing_high.iloc[i]):
-                sh_prices.append(swing_high.iloc[i])
-                if len(sh_prices) > 4:
-                    sh_prices.pop(0)
-
-            if pd.notna(swing_low.iloc[i]):
-                sl_prices.append(swing_low.iloc[i])
-                if len(sl_prices) > 4:
-                    sl_prices.pop(0)
-
-            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
-                if sh_prices[-1] > sh_prices[-2] and sl_prices[-1] > sl_prices[-2]:
-                    trend_up_arr[i] = True
-                elif sh_prices[-1] < sh_prices[-2] and sl_prices[-1] < sl_prices[-2]:
-                    trend_down_arr[i] = True
-                elif i > 0:
-                    trend_up_arr[i] = trend_up_arr[i - 1]
-                    trend_down_arr[i] = trend_down_arr[i - 1]
-            elif i > 0:
-                trend_up_arr[i] = trend_up_arr[i - 1]
-                trend_down_arr[i] = trend_down_arr[i - 1]
-
-            if sl_prices:
-                nearest_support[i] = sl_prices[-1]
-            if sh_prices:
-                nearest_resistance[i] = sh_prices[-1]
-
-            c = close.iloc[i]
-            if not np.isnan(nearest_support[i]) and not np.isnan(nearest_resistance[i]):
-                zone_range = nearest_resistance[i] - nearest_support[i]
-                if zone_range > 0:
-                    pos_pct = (c - nearest_support[i]) / zone_range
-                    in_demand_zone[i] = pos_pct < 0.35
-                    in_supply_zone[i] = pos_pct > 0.65
-
-        return DataFrame({
-            "trend_up": trend_up_arr,
-            "trend_down": trend_down_arr,
-            "support": nearest_support,
-            "resistance": nearest_resistance,
-            "in_demand": in_demand_zone,
-            "in_supply": in_supply_zone,
-        }, index=high.index)
-
-    # =====================
-    # 工具：K线形态检测
-    # =====================
-
-    @staticmethod
-    def _detect_candle_patterns(
-        open_: pd.Series,
-        high: pd.Series,
-        low: pd.Series,
-        close: pd.Series,
-        pin_bar_wick_ratio: float = 0.6,
-    ) -> tuple[pd.Series, pd.Series, pd.Series, pd.Series]:
-        body = (close - open_).abs()
-        total_range = (high - low).replace(0, 0.0001)
-
-        upper_wick = high - close.where(close > open_, open_)
-        lower_wick = open_.where(close > open_, close) - low
-        is_pin = (upper_wick + lower_wick) / total_range > pin_bar_wick_ratio
-
-        bullish_pin = is_pin & (close > open_) & (lower_wick > upper_wick)
-        bearish_pin = is_pin & (close < open_) & (upper_wick > lower_wick)
-
-        prev_open = open_.shift(1)
-        prev_close = close.shift(1)
-        bullish_engulf = (close > prev_open) & (open_ < prev_close) & (close > open_)
-        bearish_engulf = (close < prev_open) & (open_ > prev_close) & (close < open_)
-
-        return bullish_pin, bearish_pin, bullish_engulf, bearish_engulf
-
-    # =====================
-    # 工具：冷却期正确实现（修复 bug）
-    # =====================
-
-    def _apply_cooldown(self, signal: pd.Series, cooldown_bars: int) -> pd.Series:
-        """
-        正确应用冷却期：入场后才冷却，而非条件满足就冷却。
-        
-        原逻辑 bug：long_base.rolling(cooldown).max().shift(1) == 0
-        - 当市场持续满足入场条件时，rolling window 里永远有 True
-        - 导致冷却期无限阻止下单
-        
-        修复逻辑：遍历 K 线，模拟"入场 -> 冷却"过程。
-        - 满足条件 + 距离上次入场 > cooldown -> 允许入场
-        - 入场后 cooldown 根 K 线内不再入场
-        """
-        n = len(signal)
-        result = [False] * n
-        last_entry = -99999  # 上次入场的 bar 索引
-        
-        # 遍历（对 numpy array 操作，O(n) 约几毫秒）
-        values = signal.values  # numpy array，快速访问
-        for i in range(n):
-            if values[i] and (i - last_entry) > cooldown_bars:
-                result[i] = True
-                last_entry = i
-        
-        return pd.Series(result, index=signal.index)
+    # ── 保护性止损 ──
+    stoploss = -0.10  # 硬止损 10%
 
     # ================================================================
-    #  信息时间框架 — D1 宏观结构
+    #  杠杆
     # ================================================================
 
-    @informative("1d")
-    def populate_indicators_1d(
+    def leverage(
+        self, pair: str, current_time: datetime, current_rate: float,
+        proposed_leverage: float, max_leverage: float, side: str,
+        **kwargs,
+    ) -> float:
+        """20x 杠杆起步，验证胜率后再上量"""
+        return min(20.0, max_leverage)
+
+    # ================================================================
+    #  信息时间框架 — 15m 趋势判断 + S/R
+    # ================================================================
+
+    @informative("15m")
+    def populate_indicators_15m(
         self, dataframe: DataFrame, metadata: dict
     ) -> DataFrame:
-        sh, sl = self._detect_swing_points(
-            dataframe["high"], dataframe["low"],
-            self.swing_lookback_d1.value,
-        )
-        structure = self._build_structure(
-            dataframe["high"], dataframe["low"], dataframe["close"],
-            sh, sl,
-        )
-        dataframe["trend_up"] = structure["trend_up"]
-        dataframe["trend_down"] = structure["trend_down"]
-        return dataframe
+        """15m级别：EMA趋势方向 + swing point S/R。"""
 
-    # ================================================================
-    #  信息时间框架 — 4H 趋势强度（原版保留）
-    # ================================================================
+        # ── EMA 趋势方向 ──
+        ema_period = self.trend_ema_period.value
+        dataframe["ema_fast"] = dataframe["close"].ewm(span=ema_period, adjust=False).mean()
+        dataframe["ema_slow"] = dataframe["close"].ewm(span=ema_period * 2.5, adjust=False).mean()
 
-    @informative("4h")
-    def populate_indicators_4h(
-        self, dataframe: DataFrame, metadata: dict
-    ) -> DataFrame:
-        sh, sl = self._detect_swing_points(
-            dataframe["high"], dataframe["low"],
-            self.swing_lookback_h4.value,
-        )
-        structure = self._build_structure(
-            dataframe["high"], dataframe["low"], dataframe["close"],
-            sh, sl,
+        dataframe["trend_up"] = dataframe["ema_fast"] > dataframe["ema_slow"]
+        dataframe["trend_down"] = dataframe["ema_fast"] < dataframe["ema_slow"]
+
+        # ── Swing Point 支撑/阻力 ──
+        high = dataframe["high"].tolist()
+        low = dataframe["low"].tolist()
+        close = dataframe["close"].tolist()
+
+        sh, sl = self._detect_swing_points(high, low, window=5)
+        trend_up_arr, trend_down_arr, support_arr, resistance_arr = self._build_structure(
+            high, low, close, sh, sl,
         )
 
-        # 趋势强度计算（原版逻辑）
-        sh_prices = []
-        sl_prices = []
-        trend_strength_up = np.full(len(dataframe), np.nan)
-        trend_strength_down = np.full(len(dataframe), np.nan)
-
-        for i in range(len(dataframe)):
-            if pd.notna(sh.iloc[i]):
-                sh_prices.append(sh.iloc[i])
-                if len(sh_prices) > 4:
-                    sh_prices.pop(0)
-            if pd.notna(sl.iloc[i]):
-                sl_prices.append(sl.iloc[i])
-                if len(sl_prices) > 4:
-                    sl_prices.pop(0)
-
-            if len(sh_prices) >= 2 and len(sl_prices) >= 2:
-                hh_dist = (sh_prices[-1] - sh_prices[-2]) / sh_prices[-2] if sh_prices[-2] > 0 else 0
-                hl_dist = (sl_prices[-1] - sl_prices[-2]) / sl_prices[-2] if sl_prices[-2] > 0 else 0
-                trend_strength_up[i] = hh_dist + hl_dist
-                trend_strength_down[i] = -(hh_dist + hl_dist)
-
-        dataframe["trend_strength_up"] = pd.Series(trend_strength_up, index=dataframe.index)
-        dataframe["trend_strength_down"] = pd.Series(trend_strength_down, index=dataframe.index)
-
-        min_strength = self.trend_strength_min.value / 100.0
-        dataframe["strong_uptrend"] = dataframe["trend_strength_up"] > min_strength
-        dataframe["strong_downtrend"] = dataframe["trend_strength_down"] > min_strength
+        dataframe["trend_up_sp"] = trend_up_arr
+        dataframe["trend_down_sp"] = trend_down_arr
+        # EMA平滑S/R（避免跳变）
+        dataframe["support"] = self._ema_smooth(support_arr, alpha=0.3)
+        dataframe["resistance"] = self._ema_smooth(resistance_arr, alpha=0.3)
 
         return dataframe
 
     # ================================================================
-    #  主时间框架 — 1H 指标（含 1H S/R + 活支撑/阻力）
+    #  主框架 — 5m 级别指标
     # ================================================================
 
     def populate_indicators(
         self, dataframe: DataFrame, metadata: dict
     ) -> DataFrame:
+        """5m级别：ATR + K线形态 + EMA趋势整合。"""
+
+        # ── ATR(14) ──
+        high = dataframe["high"]
+        low = dataframe["low"]
+        close = dataframe["close"]
+        prev_close = close.shift(1)
+        tr = pd.concat([
+            high - low,
+            (high - prev_close).abs(),
+            (low - prev_close).abs(),
+        ], axis=1).max(axis=1)
+        dataframe["atr"] = tr.rolling(14).mean()
+        atr_mean = dataframe["atr"].mean()
+        dataframe["atr"] = dataframe["atr"].fillna(atr_mean)
+
         # ── K线形态 ──
         bullish_pin, bearish_pin, bullish_engulf, bearish_engulf = (
             self._detect_candle_patterns(
-                dataframe["open"],
-                dataframe["high"],
-                dataframe["low"],
-                dataframe["close"],
-                self.pin_bar_wick_ratio.value / 100.0,
+                dataframe["open"], dataframe["high"], dataframe["low"],
+                dataframe["close"], self.pin_bar_wick_ratio.value / 100.0,
             )
         )
-        dataframe["bullish_pinbar"] = bullish_pin
-        dataframe["bearish_pinbar"] = bearish_pin
-        dataframe["bullish_engulfing"] = bullish_engulf
-        dataframe["bearish_engulfing"] = bearish_engulf
         dataframe["bullish_signal"] = bullish_pin | bullish_engulf
         dataframe["bearish_signal"] = bearish_pin | bearish_engulf
 
-        # ── 1H级别 Swing Point + 结构（替代原4H S/R） ──
-        sh_1h, sl_1h = self._detect_swing_points(
-            dataframe["high"], dataframe["low"],
-            self.swing_lookback_1h.value,
-        )
-        structure_1h = self._build_structure(
-            dataframe["high"], dataframe["low"], dataframe["close"],
-            sh_1h, sl_1h,
-        )
-        dataframe["trend_up_1h"] = structure_1h["trend_up"]
-        dataframe["trend_down_1h"] = structure_1h["trend_down"]
-        dataframe["support"] = structure_1h["support"]
-        dataframe["resistance"] = structure_1h["resistance"]
-        dataframe["in_demand"] = structure_1h["in_demand"]
-        dataframe["in_supply"] = structure_1h["in_supply"]
+        # ── 5m EMA（用于短期拉回确认） ──
+        dataframe["ema5"] = close.ewm(span=5, adjust=False).mean()
+        dataframe["ema8"] = close.ewm(span=8, adjust=False).mean()
 
-        # ── 1H 活支撑/阻力检查 ──
-        touched_support = (
-            (dataframe["low"] <= dataframe["support"] * 1.005) &
-            (dataframe["low"] >= dataframe["support"] * 0.995)
-        )
-        held_support = dataframe["close"] > dataframe["support"]
-        support_tested_and_held = touched_support & held_support
-        dataframe["support_alive"] = support_tested_and_held.rolling(3, min_periods=1).max() > 0
-
-        touched_resistance = (
-            (dataframe["high"] >= dataframe["resistance"] * 0.995) &
-            (dataframe["high"] <= dataframe["resistance"] * 1.005)
-        )
-        held_resistance = dataframe["close"] < dataframe["resistance"]
-        resistance_tested_and_held = touched_resistance & held_resistance
-        dataframe["resistance_alive"] = resistance_tested_and_held.rolling(3, min_periods=1).max() > 0
-
-        # ── NaN 安全处理 ──
-        bool_cols = [
-            "trend_up_1d", "trend_down_1d",
-            "trend_up_4h", "trend_down_4h",
-            "in_demand", "in_supply",
-            "support_alive", "resistance_alive",
-            "strong_uptrend_4h", "strong_downtrend_4h",
-            "bullish_signal", "bearish_signal",
-        ]
-        for col in bool_cols:
-            if col in dataframe.columns:
+        # ── 布尔列NaN填充 ──
+        for col in ["bullish_signal", "bearish_signal"]:
             dataframe[col] = dataframe[col].fillna(False)
 
         return dataframe
 
-    # =====================
-    # 入场信号（修复冷却期逻辑）
-    # =====================
+    # ================================================================
+    #  入场逻辑
+    # ================================================================
 
-    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
-        max_dist = self.max_stop_dist.value / 100.0
+    def populate_entry_trend(
+        self, dataframe: DataFrame, metadata: dict
+    ) -> DataFrame:
+        """
+        入场逻辑。
+
+        只做顺趋势回调入场，不做S/R反向交易：
+
+        做多条件：
+        1. 15m 上升趋势（EMA_fast > EMA_slow）
+        2. 5m 价格回调到15m EMA_fast 或 支撑位附近
+        3. 5m K线止跌信号（pinbar/engulfing）
+
+        做空条件（对称）：
+        1. 15m 下降趋势
+        2. 5m 价格反弹到15m EMA_fast 或 阻力位附近
+        3. 5m K线止涨信号
+        """
         cooldown = self.cooldown_bars.value
+        dev = self.pullback_deviation.value / 100.0  # 0.5% → 0.005
 
-        bool_cols = [
-            "trend_up_1d", "trend_down_1d",
-            "trend_up_4h", "trend_down_4h",
-            "in_demand", "in_supply",
-            "support_alive", "resistance_alive",
-            "strong_uptrend_4h", "strong_downtrend_4h",
-            "bullish_signal", "bearish_signal",
+        # ── 必要列检查 ──
+        required = [
+            "ema_fast_15m", "trend_up_15m", "trend_down_15m",
+            "support_15m", "resistance_15m",
         ]
-        for col in bool_cols:
+        for col in required:
+            if col not in dataframe.columns:
+                return dataframe
+
+        # ── 布尔列填充 ──
+        for col in [
+            "bullish_signal", "bearish_signal",
+            "trend_up_15m", "trend_down_15m",
+        ]:
             if col in dataframe.columns:
                 dataframe[col] = dataframe[col].fillna(False)
 
-        # ── 做多（使用1H S/R） ──
-        long_stop_dist = (dataframe["open"] - dataframe["support"]) / dataframe["open"]
+        # ═══════════════════════════════════════════════════════════
+        #  做多：上升趋势 + 回调到EMA/支撑 + 止跌信号
+        # ═══════════════════════════════════════════════════════════
 
-        long_base = (
-            dataframe["trend_up_1d"]
-            & dataframe["in_demand"]
-            & (long_stop_dist <= max_dist)
-            & (long_stop_dist > 0.003)
-            & dataframe["support_alive"]
-            & dataframe["strong_uptrend_4h"]
+        # 条件1：15m 上升趋势
+        trend_up = dataframe["trend_up_15m"]
+
+        # 条件2：价格在EMA20或支撑位附近（回调到顺趋势的支撑区）
+        near_ema = (
+            (dataframe["low"] <= dataframe["ema_fast_15m"] * (1.0 + dev * 0.5)) &
+            (dataframe["low"] >= dataframe["ema_fast_15m"] * (1.0 - dev * 2.0))
         )
-
-        # ✅ 修复：正确应用冷却期（基于实际入场，而非条件满足）
-        long_entries = self._apply_cooldown(long_base, cooldown)
-        dataframe.loc[long_entries, "enter_long"] = 1
-
-        # ── 做空（使用1H S/R） ──
-        short_stop_dist = (dataframe["resistance"] - dataframe["open"]) / dataframe["open"]
-
-        short_base = (
-            dataframe["trend_down_1d"]
-            & dataframe["in_supply"]
-            & (short_stop_dist <= max_dist)
-            & (short_stop_dist > 0.003)
-            & dataframe["resistance_alive"]
-            & dataframe["strong_downtrend_4h"]
+        near_support = (
+            (dataframe["low"] <= dataframe["support_15m"] * (1.0 + dev)) &
+            (dataframe["low"] >= dataframe["support_15m"] * (1.0 - dev))
         )
+        pullback_long = near_ema | near_support
 
-        # ✅ 修复：正确应用冷却期（基于实际入场，而非条件满足）
-        short_entries = self._apply_cooldown(short_base, cooldown)
-        dataframe.loc[short_entries, "enter_short"] = 1
+        # 条件3：K线止跌信号
+        signal_long = dataframe["bullish_signal"]
+
+        # 综合入场
+        enter_long = trend_up & pullback_long & signal_long
+        long_recent = enter_long.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[enter_long & long_recent, "enter_long"] = 1
+
+        # ═══════════════════════════════════════════════════════════
+        #  做空：下降趋势 + 反弹到EMA/阻力 + 止涨信号
+        # ═══════════════════════════════════════════════════════════
+
+        # 条件1：15m 下降趋势
+        trend_down = dataframe["trend_down_15m"]
+
+        # 条件2：价格在EMA20或阻力位附近（反弹到顺趋势的阻力区）
+        near_ema_short = (
+            (dataframe["high"] >= dataframe["ema_fast_15m"] * (1.0 - dev * 0.5)) &
+            (dataframe["high"] <= dataframe["ema_fast_15m"] * (1.0 + dev * 2.0))
+        )
+        near_resistance = (
+            (dataframe["high"] >= dataframe["resistance_15m"] * (1.0 - dev)) &
+            (dataframe["high"] <= dataframe["resistance_15m"] * (1.0 + dev))
+        )
+        pullback_short = near_ema_short | near_resistance
+
+        # 条件3：K线止涨信号
+        signal_short = dataframe["bearish_signal"]
+
+        # 综合入场
+        enter_short = trend_down & pullback_short & signal_short
+        short_recent = enter_short.rolling(cooldown, min_periods=1).max().shift(1) == 0
+        dataframe.loc[enter_short & short_recent, "enter_short"] = 1
 
         return dataframe
 
-    # =====================
-    # 出场信号
-    # =====================
+    # ================================================================
+    #  exit_trend（freqtrade 2025.11 强制要求，即使 use_exit_signal=False）
+    # ================================================================
 
     def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
-        exit_long = ~dataframe["trend_up_1d"].fillna(True)
-        dataframe.loc[exit_long, "exit_long"] = 1
-
-        exit_short = dataframe["trend_up_1d"].fillna(False)
-        dataframe.loc[exit_short, "exit_short"] = 1
-
+        """出场完全由 custom_stoploss + custom_exit 管理。"""
         return dataframe
 
-    # =====================
-    # 动态止损（基于1H S/R）
-    # =====================
+    # ================================================================
+    #  出场 — 止损（ATR动态）
+    # ================================================================
 
     def custom_stoploss(
         self,
@@ -404,48 +276,240 @@ class StructureFlowStrategyV22d(IStrategy):
         after_fill: bool,
         **kwargs,
     ) -> float:
+        """
+        止损 = 入场价 ± ATR × atr_mult_stop
+
+        - ATR值从入场K线锁定，持仓期间不变
+        - 做多：entry_price - (locked_atr × mult)
+        - 做空：entry_price + (locked_atr × mult)
+        - 配20x杠杆，ATR×1.0 ≈ 对应约 $3.7 止损（当前5m ATR~$3.74）
+        """
         dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
         if dataframe is None or len(dataframe) == 0:
             return -0.02 if not trade.is_short else 0.02
 
-        last = dataframe.iloc[-1]
+        entry_row = self._get_entry_row(dataframe, trade)
+        if entry_row is None:
+            return -0.02 if not trade.is_short else 0.02
+
+        atr = entry_row.get("atr", np.nan)
+        if pd.isna(atr) or atr <= 0:
+            return -0.02 if not trade.is_short else 0.02
+
+        mult = self.atr_mult_stop.value
 
         if not trade.is_short:
-            support = last.get("support", np.nan)
-            if pd.isna(support) or support <= 0:
-                return -0.02
-            sl_price = support * 0.999
-            sl_ratio = (sl_price / current_rate) - 1.0
-            return max(sl_ratio, -0.15)
+            sl_price = trade.open_rate - (atr * mult)
+            sl_ratio = (sl_price / trade.open_rate) - 1.0
+            return max(sl_ratio, -self.stoploss)
         else:
-            resistance = last.get("resistance", np.nan)
-            if pd.isna(resistance) or resistance <= 0:
-                return 0.02
-            sl_price = resistance * 1.001
-            sl_ratio = 1.0 - (sl_price / current_rate)
-            return min(sl_ratio, 0.15)
+            sl_price = trade.open_rate + (atr * mult)
+            sl_ratio = 1.0 - (sl_price / trade.open_rate)
+            return min(sl_ratio, self.stoploss)
 
-    # =====================
-    # Plot config
-    # =====================
+    # ================================================================
+    #  出场 — 止盈（ATR动态）+ 时间止损
+    # ================================================================
 
-    @staticmethod
-    def plot_config() -> dict:
-        return {
-            "main_plot": {
-                "support": {"color": "green", "type": "line"},
-                "resistance": {"color": "red", "type": "line"},
-            },
-            "subplots": {
-                "signals": {
-                    "bullish_pinbar": {"color": "green", "type": "scatter"},
-                    "bearish_pinbar": {"color": "red", "type": "scatter"},
-                },
-                "filters": {
-                    "support_alive": {"color": "green", "type": "line"},
-                    "resistance_alive": {"color": "red", "type": "line"},
-                    "strong_uptrend_4h": {"color": "blue", "type": "line"},
-                    "strong_downtrend_4h": {"color": "orange", "type": "line"},
-                },
-            },
-        }
+    def custom_exit(
+        self,
+        pair: str,
+        trade: Trade,
+        current_time: datetime,
+        current_rate: float,
+        current_profit: float,
+        **kwargs,
+    ) -> str | None:
+        """
+        出场逻辑：
+        1. ATR止盈：利润达到入场时锁定的 ATR × atr_mult_tp → 止盈
+        2. 时间止损：持仓超过 time_stop_minutes → 强制出场
+        """
+        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
+        if dataframe is None or len(dataframe) == 0:
+            return None
+
+        entry_row = self._get_entry_row(dataframe, trade)
+        if entry_row is None:
+            return None
+
+        atr = entry_row.get("atr", np.nan)
+        if pd.isna(atr) or atr <= 0:
+            return None
+
+        # 1. ATR 止盈
+        tp_mult = self.atr_mult_tp.value
+        tp_ratio = (atr * tp_mult) / trade.open_rate
+
+        if current_profit >= tp_ratio:
+            return "atr_tp"
+
+        # 2. 时间止损
+        elapsed = (current_time - trade.open_date).total_seconds() / 60.0
+        if elapsed >= self.time_stop_minutes:
+            return "time_stop"
+
+        return None
+
+    # ================================================================
+    #  工具函数
+    # ================================================================
+
+    def _detect_swing_points(
+        self, highs: list, lows: list, window: int = 5
+    ):
+        """
+        Swing High / Swing Low 检测。
+
+        当一根K线的最高价高于其两侧window根K线的最高价时，标记为Swing High。
+        Swing Low同理。
+        """
+        n = len(highs)
+        swing_high = [np.nan] * n
+        swing_low = [np.nan] * n
+
+        for i in range(window, n - window):
+            # Swing High
+            is_high = True
+            for j in range(i - window, i + window + 1):
+                if j == i:
+                    continue
+                if highs[j] >= highs[i]:
+                    is_high = False
+                    break
+            if is_high:
+                swing_high[i] = highs[i]
+
+            # Swing Low
+            is_low = True
+            for j in range(i - window, i + window + 1):
+                if j == i:
+                    continue
+                if lows[j] <= lows[i]:
+                    is_low = False
+                    break
+            if is_low:
+                swing_low[i] = lows[i]
+
+        return swing_high, swing_low
+
+    def _build_structure(
+        self, highs: list, lows: list, closes: list,
+        swing_high: list, swing_low: list,
+    ):
+        """构建趋势结构和支撑/阻力位。"""
+        n = len(highs)
+        trend_up = [False] * n
+        trend_down = [False] * n
+        support = [np.nan] * n
+        resistance = [np.nan] * n
+
+        # 用最近4个swing point的位置判断
+        last_sh_idx = -1
+        last_sl_idx = -1
+        prev_sh = []
+        prev_sl = []
+
+        for i in range(n):
+            if not np.isnan(swing_high[i]):
+                prev_sh.append(swing_high[i])
+                last_sh_idx = i
+                if len(prev_sh) > 4:
+                    prev_sh.pop(0)
+
+            if not np.isnan(swing_low[i]):
+                prev_sl.append(swing_low[i])
+                last_sl_idx = i
+                if len(prev_sl) > 4:
+                    prev_sl.pop(0)
+
+            # 趋势判断：最新的HH > 次新的HH = 上升趋势中的higher high
+            if len(prev_sh) >= 2 and prev_sh[-1] > prev_sh[-2]:
+                trend_up[i] = True
+
+            # 趋势判断：最新的LL < 次新的LL = 下降趋势中的lower low
+            if len(prev_sl) >= 2 and prev_sl[-1] < prev_sl[-2]:
+                trend_down[i] = True
+
+            # 支撑 = 最近的有效Swing Low（EMA平滑后在调用侧处理）
+            if prev_sl:
+                support[i] = prev_sl[-1]
+            if prev_sh:
+                resistance[i] = prev_sh[-1]
+
+        return trend_up, trend_down, support, resistance
+
+    def _ema_smooth(self, values: list, alpha: float = 0.3):
+        """对数组做EMA平滑，避免跳变。"""
+        result = [np.nan] * len(values)
+        ema = None
+        for i, v in enumerate(values):
+            if pd.isna(v) or v is None:
+                if ema is not None:
+                    result[i] = ema
+                continue
+            if ema is None:
+                ema = v
+            else:
+                ema = alpha * v + (1 - alpha) * ema
+            result[i] = ema
+        return np.array(result)
+
+    def _detect_candle_patterns(
+        self, opens, highs, lows, closes, wick_ratio=0.6,
+    ):
+        """检测K线形态：pinbar（锤子线/射击星）和吞没形态。"""
+        n = len(opens)
+        bullish_pin = [False] * n
+        bearish_pin = [False] * n
+        bullish_engulf = [False] * n
+        bearish_engulf = [False] * n
+
+        for i in range(n):
+            o, h, l, c = opens[i], highs[i], lows[i], closes[i]
+            total_range = h - l if h > l else 0.001
+
+            is_bullish = c > o
+            is_bearish = c < o
+
+            body = abs(c - o)
+            upper_wick = h - max(c, o)
+            lower_wick = min(c, o) - l
+
+            # Pinbar：影线 > total_range × wick_ratio
+            if is_bullish and lower_wick / total_range > wick_ratio:
+                bullish_pin[i] = True
+            if is_bearish and upper_wick / total_range > wick_ratio:
+                bearish_pin[i] = True
+
+            # 吞没形态
+            if i > 0:
+                prev_o = opens[i - 1]
+                prev_c = closes[i - 1]
+                if is_bullish and c > prev_o and o < prev_c:
+                    bullish_engulf[i] = True
+                if is_bearish and c < prev_o and o > prev_c:
+                    bearish_engulf[i] = True
+
+        return (
+            pd.Series(bullish_pin),
+            pd.Series(bearish_pin),
+            pd.Series(bullish_engulf),
+            pd.Series(bearish_engulf),
+        )
+
+    def _get_entry_row(self, dataframe: DataFrame, trade: Trade):
+        """查找入场K线行，兼容live/backtesting两种模式。"""
+        if "date" in dataframe.columns:
+            entry_mask = pd.to_datetime(dataframe["date"]) <= trade.open_date
+            if not entry_mask.any():
+                return None
+            return dataframe[entry_mask].iloc[-1]
+        else:
+            try:
+                idx = dataframe.index.get_indexer([trade.open_date], method="pad")
+                if idx[0] < 0 or idx[0] >= len(dataframe):
+                    return None
+                return dataframe.iloc[idx[0]]
+            except (TypeError, ValueError):
+                return None