from freqtrade.strategy import IStrategy from pandas import DataFrame import talib.abstract as ta import numpy as np import logging logger = logging.getLogger(__name__) class FreqAIStrategy(IStrategy): """FreqAI 趋势捕捉策略 - 优化版""" INTERFACE_VERSION = 3 timeframe = '1h' stoploss = -0.03 # 收紧止损到 3% # 移动止损：盈利 5% 后启用 2% 回撤止损 (更宽松) trailing_stop = True trailing_stop_positive = 0.02 trailing_stop_positive_offset = 0.05 trailing_only_offset_is_reached = True # FreqAI 必需配置 process_only_new_candles = True startup_candle_count = 300 # 增加启动 K 线数量以计算更长周期指标 def feature_engineering_expand_all(self, dataframe: DataFrame, period: int, metadata: dict, **kwargs) -> DataFrame: """ 高级特征工程 - FreqAI 自动扩展 特征必须以 % 为前缀 """ # 动量指标 dataframe["%rsi"] = ta.RSI(dataframe, timeperiod=period) dataframe["%mfi"] = ta.MFI(dataframe, timeperiod=period) dataframe["%adx"] = ta.ADX(dataframe, timeperiod=period) dataframe["%cci"] = ta.CCI(dataframe, timeperiod=period) dataframe["%mom"] = ta.MOM(dataframe, timeperiod=period) dataframe["%roc"] = ta.ROC(dataframe, timeperiod=period) # Bollinger Bands bollinger = ta.BBANDS(dataframe, timeperiod=period) dataframe["%bb_width"] = (bollinger['upperband'] - bollinger['lowerband']) / dataframe['close'] dataframe["%bb_pct"] = (dataframe['close'] - bollinger['lowerband']) / \ (bollinger['upperband'] - bollinger['lowerband'] + 1e-10) # ATR 波动率 dataframe["%atr"] = ta.ATR(dataframe, timeperiod=period) / dataframe['close'] return dataframe def feature_engineering_standard(self, dataframe: DataFrame, **kwargs) -> DataFrame: """ 标准特征工程 - 趋势特征增强版 """ # ===== 趋势指标 ===== # EMA 均线系统 dataframe["%ema_12"] = ta.EMA(dataframe, timeperiod=12) / dataframe['close'] - 1 dataframe["%ema_26"] = ta.EMA(dataframe, timeperiod=26) / dataframe['close'] - 1 dataframe["%ema_50"] = ta.EMA(dataframe, timeperiod=50) / dataframe['close'] - 1 dataframe["%ema_200"] = ta.EMA(dataframe, timeperiod=200) / dataframe['close'] - 1 # EMA 交叉信号 ema_12 = ta.EMA(dataframe, timeperiod=12) ema_26 = ta.EMA(dataframe, timeperiod=26) ema_50 = ta.EMA(dataframe, timeperiod=50) dataframe["%ema_12_26_cross"] = (ema_12 - ema_26) / dataframe['close'] dataframe["%ema_26_50_cross"] = (ema_26 - ema_50) / dataframe['close'] # ===== MACD ===== macd = ta.MACD(dataframe) dataframe["%macd"] = macd["macd"] / dataframe['close'] dataframe["%macdsignal"] = macd["macdsignal"] / dataframe['close'] dataframe["%macdhist"] = macd["macdhist"] / dataframe['close'] # ===== ADX 趋势强度 ===== dataframe["%adx_14"] = ta.ADX(dataframe, timeperiod=14) dataframe["%plus_di"] = ta.PLUS_DI(dataframe, timeperiod=14) dataframe["%minus_di"] = ta.MINUS_DI(dataframe, timeperiod=14) dataframe["%di_diff"] = (dataframe["%plus_di"] - dataframe["%minus_di"]) / 100 # ===== RSI ===== dataframe["%rsi_14"] = ta.RSI(dataframe, timeperiod=14) dataframe["%rsi_7"] = ta.RSI(dataframe, timeperiod=7) # ===== 价格动量 ===== dataframe["%return_1h"] = dataframe["close"].pct_change(1) dataframe["%return_4h"] = dataframe["close"].pct_change(4) dataframe["%return_12h"] = dataframe["close"].pct_change(12) dataframe["%return_24h"] = dataframe["close"].pct_change(24) # ===== 波动率 ===== dataframe["%volatility_12h"] = dataframe["close"].pct_change().rolling(12).std() dataframe["%volatility_24h"] = dataframe["close"].pct_change().rolling(24).std() # ===== 成交量 ===== dataframe["%volume_pct_change"] = dataframe["volume"].pct_change() dataframe["%volume_ma_ratio"] = dataframe["volume"] / dataframe["volume"].rolling(20).mean() # ===== 时间特征 ===== dataframe["%day_of_week"] = dataframe["date"].dt.dayofweek dataframe["%hour_of_day"] = dataframe["date"].dt.hour # ===== 趋势状态 (用于回测辅助) ===== # 保存 ADX 和 EMA 状态供入场过滤使用 dataframe["adx_value"] = ta.ADX(dataframe, timeperiod=14) dataframe["ema_bullish"] = (ema_12 > ema_26).astype(int) return dataframe def set_freqai_targets(self, dataframe: DataFrame, metadata: dict, **kwargs) -> DataFrame: """ 设置预测目标 - 延长到 12 小时 (12 个 1h K 线) """ # 预测未来 12 个周期的收益率 dataframe["&-s_return"] = ( dataframe["close"].shift(-12) / dataframe["close"] - 1 ) return dataframe def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame: # FreqAI 启动入口 dataframe = self.freqai.start(dataframe, metadata, self) return dataframe def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: # 确保有预测结果 if '&-s_return_mean' in dataframe.columns: # Debug: 打印预测值范围 pred_mean = dataframe['&-s_return_mean'].dropna() pred_std = dataframe['&-s_return_std'].dropna() if '&-s_return_std' in dataframe.columns else None if len(pred_mean) > 0: logger.info(f"[DEBUG] Predictions for {metadata['pair']}: mean range [{pred_mean.min():.4f}, {pred_mean.max():.4f}], avg={pred_mean.mean():.4f}") if pred_std is not None and len(pred_std) > 0: logger.info(f"[DEBUG] Prediction std range [{pred_std.min():.4f}, {pred_std.max():.4f}]") # ===== 纯趋势跟踪入场 ===== # 条件： # 1. 预测为正 (任意正收益) # 2. ADX > 25 (趋势强劲) # 3. EMA 多头排列 (12 > 26) dataframe.loc[ ( (dataframe['&-s_return_mean'] > 0) & # 预测为正 (无阈值) (dataframe['adx_value'] > 25) & # 趋势强劲 (dataframe['ema_bullish'] == 1) & # EMA 多头 (dataframe['volume'] > 0) ), 'enter_long'] = 1 # 移除激进入场 - 只保留趋势确认入场 return dataframe def populate_exit_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame: if '&-s_return_mean' in dataframe.columns: # ===== 预测转空出场 ===== dataframe.loc[ ( (dataframe['&-s_return_mean'] < -0.003) | # 预测下跌 > 0.3% ( (dataframe['&-s_return_mean'] < 0) & # 预测转负 (dataframe['ema_bullish'] == 0) # 且 EMA 转空头 ) ), 'exit_long'] = 1 return dataframe