优化！

backend/app/crypto_agent/crypto_agent.py

@@ -509,6 +509,7 @@ class CryptoAgent:
                     await self._check_and_set_pending_tp_sl_hyperliquid()
                 if self.bitget:
                     await self._check_and_set_pending_tp_sl_bitget()
+                    await self._check_bitget_missing_tp_sl()  # 兜底：检查缺少的 TP/SL 并补救
 
                 for symbol in self.symbols:
                     await self.analyze_symbol(symbol)
@@ -721,13 +722,13 @@ class CryptoAgent:
             if self.settings.paper_trading_enabled:
                 logger.info(f"\n📊 【模拟盘】")
                 paper_positions, paper_account, paper_pending = self._get_paper_trading_state()
-                paper_signal = self._select_signal_for_platform(valid_signals, 'PaperTrading')
+                paper_signal = self._select_signal_for_platform(valid_signals, 'PaperTrading', market_state=market_signal.get('market_state', '中性'))
                 if paper_signal:
                     logger.info(
                         f"   采用信号: {paper_signal.get('timeframe', 'unknown')} | "
                         f"{paper_signal.get('action')} | {paper_signal.get('confidence', 0)}%"
                     )
-                    trading_signal = self._build_execution_signal(symbol, paper_signal, current_price)
+                    trading_signal = self._build_execution_signal(symbol, paper_signal, current_price, market_signal)
                     paper_decision = self.execute_signal_with_rules(
                         trading_signal, 'PaperTrading', paper_account, paper_positions, paper_pending
                     )
@@ -745,13 +746,13 @@ class CryptoAgent:
             if self.hyperliquid:
                 logger.info(f"\n🔥 【Hyperliquid】")
                 hl_positions, hl_account, hl_pending = self._get_hyperliquid_trading_state()
-                hl_signal = self._select_signal_for_platform(valid_signals, 'Hyperliquid')
+                hl_signal = self._select_signal_for_platform(valid_signals, 'Hyperliquid', market_state=market_signal.get('market_state', '中性'))
                 if hl_signal:
                     logger.info(
                         f"   采用信号: {hl_signal.get('timeframe', 'unknown')} | "
                         f"{hl_signal.get('action')} | {hl_signal.get('confidence', 0)}%"
                     )
-                    trading_signal = self._build_execution_signal(symbol, hl_signal, current_price)
+                    trading_signal = self._build_execution_signal(symbol, hl_signal, current_price, market_signal)
                     hl_decision = self.execute_signal_with_rules(
                         trading_signal, 'Hyperliquid', hl_account, hl_positions, hl_pending
                     )
@@ -769,13 +770,13 @@ class CryptoAgent:
             if self.bitget:
                 logger.info(f"\n🔥 【Bitget】")
                 bg_positions, bg_account, bg_pending = self._get_bitget_trading_state()
-                bg_signal = self._select_signal_for_platform(valid_signals, 'Bitget')
+                bg_signal = self._select_signal_for_platform(valid_signals, 'Bitget', market_state=market_signal.get('market_state', '中性'))
                 if bg_signal:
                     logger.info(
                         f"   采用信号: {bg_signal.get('timeframe', 'unknown')} | "
                         f"{bg_signal.get('action')} | {bg_signal.get('confidence', 0)}%"
                     )
-                    trading_signal = self._build_execution_signal(symbol, bg_signal, current_price)
+                    trading_signal = self._build_execution_signal(symbol, bg_signal, current_price, market_signal)
                     bg_decision = self.execute_signal_with_rules(
                         trading_signal, 'Bitget', bg_account, bg_positions, bg_pending
                     )
@@ -1223,25 +1224,48 @@ class CryptoAgent:
         sorted_signals = sorted(signals, key=lambda x: x.get('confidence', 0), reverse=True)
         return sorted_signals[0]
 
-    def _select_signal_for_platform(self, signals: List[Dict[str, Any]], platform_name: str) -> Optional[Dict[str, Any]]:
-        """根据平台偏好选择最适合执行的信号"""
+    def _select_signal_for_platform(self, signals: List[Dict[str, Any]],
+                                      platform_name: str,
+                                      market_state: str = '中性') -> Optional[Dict[str, Any]]:
+        """根据平台偏好和市场状态选择最适合执行的信号"""
         if not signals:
             return None
 
+        # 震荡市：趋势信号降权（信心 × 0.8），优先选择日内反转信号
+        adjusted_signals = []
+        for signal in signals:
+            s = dict(signal)  # 不修改原信号
+            confidence = s.get('confidence', 50)
+            lane = s.get('timeframe') or s.get('type') or 'unknown'
+
+            if market_state == '震荡市' and lane == 'medium_term':
+                confidence = int(confidence * 0.8)
+                s['_regime_adjusted'] = True
+                s['_original_confidence'] = s.get('confidence', 50)
+                s['confidence'] = confidence
+            elif market_state in ('趋势市', '日内趋势') and lane == 'short_term':
+                pass  # 趋势市不降权日内信号
+            adjusted_signals.append(s)
+
         lane_priority = self.PLATFORM_SIGNAL_PRIORITY.get(platform_name, ['short_term', 'medium_term'])
         by_lane: Dict[str, List[Dict[str, Any]]] = {}
-        for signal in signals:
+        for signal in adjusted_signals:
             lane = signal.get('timeframe') or signal.get('type') or 'unknown'
             by_lane.setdefault(lane, []).append(signal)
 
         for lane in lane_priority:
             candidates = by_lane.get(lane, [])
             if candidates:
-                return sorted(candidates, key=lambda item: item.get('confidence', 0), reverse=True)[0]
+                best = sorted(candidates, key=lambda item: item.get('confidence', 0), reverse=True)[0]
+                if best.get('_regime_adjusted'):
+                    logger.info(f"   📊 震荡市降权: {lane} 原始信心={best.get('_original_confidence')} → 调整后={best.get('confidence')}")
+                return best
 
-        return sorted(signals, key=lambda item: item.get('confidence', 0), reverse=True)[0]
+        return sorted(adjusted_signals, key=lambda item: item.get('confidence', 0), reverse=True)[0]
 
-    def _build_execution_signal(self, symbol: str, signal: Dict[str, Any], current_price: float) -> Dict[str, Any]:
+    def _build_execution_signal(self, symbol: str, signal: Dict[str, Any],
+                                   current_price: float,
+                                   market_signal: Dict[str, Any] = None) -> Dict[str, Any]:
         """构建传给执行规则层的标准信号格式"""
         signal_type = signal.get('timeframe') or signal.get('type') or 'medium_term'
         position_size = signal.get('position_size') or self.SIGNAL_POSITION_SIZE_DEFAULTS.get(signal_type, 'light')
@@ -1259,6 +1283,9 @@ class CryptoAgent:
             'timeframe': signal_type,
             'type': signal_type,
             'position_size': position_size,
+            'market_state': market_signal.get('market_state', '中性') if market_signal else '中性',
+            'regime': (market_signal.get('range_metrics') or {}).get('regime', '') if market_signal else '',
+            'funding_rate_data': market_signal.get('funding_rate_data') if market_signal else None,
         }
 
     def _get_signal_for_decision(self, market_signal: Dict[str, Any], decision: Dict[str, Any]) -> Dict[str, Any]:
@@ -2254,6 +2281,25 @@ class CryptoAgent:
             default_positions=self.SIGNAL_POSITION_SIZE_DEFAULTS,
         )
 
+        # 市场状态仓位调整：震荡市降低仓位，避免来回被止损
+        regime = signal.get('regime', '')
+        REGIME_MARGIN_MULTIPLIERS = {
+            'ranging': 0.5,       # 震荡市：仓位减半
+            'transitional': 0.7,  # 过渡期：仓位七折
+            'weak_trend': 0.9,    # 弱趋势：仓位九折
+            'strong_trend': 1.0,  # 强趋势：满仓
+        }
+        regime_multiplier = REGIME_MARGIN_MULTIPLIERS.get(regime, 1.0)
+        if regime_multiplier < 1.0:
+            logger.info(f"   📊 市场状态调整: regime={regime}, 仓位系数={regime_multiplier}")
+            target_margin_pct *= regime_multiplier
+
+        # 连败降温：连败时降低仓位
+        streak_multiplier = signal.get('_streak_margin_multiplier', 1.0)
+        if streak_multiplier < 1.0:
+            logger.info(f"   📊 连败降温: 仓位系数={streak_multiplier}")
+            target_margin_pct *= streak_multiplier
+
         margin, _, budget_reason = calculate_margin_and_position_value(
             balance=balance,
             available_margin=available,
@@ -2386,6 +2432,53 @@ class CryptoAgent:
         # 无反向订单 → 正常开仓
         return "OPEN", "无反向订单，正常开仓"
 
+    def _check_losing_streak(self, platform_name: str, max_lookback: int = 5) -> Dict[str, Any]:
+        """
+        检查近期交易连败情况
+
+        Returns:
+            {
+                'losing_streak': int,
+                'should_cool_down': bool,
+                'margin_multiplier': float,
+                'reason': str,
+            }
+        """
+        recent_orders = []
+
+        if platform_name == 'PaperTrading' and self.paper_trading:
+            recent_orders = self.paper_trading.get_order_history(limit=max_lookback)
+        # Bitget/Hyperliquid 实盘暂不查询历史（API 限制），用空列表
+
+        if not recent_orders:
+            return {'losing_streak': 0, 'should_cool_down': False, 'margin_multiplier': 1.0, 'reason': ''}
+
+        # 计算连败（从最近的交易往回数）
+        losing_streak = 0
+        for order in recent_orders:
+            pnl = order.get('pnl_amount', 0) or 0
+            if pnl < 0:
+                losing_streak += 1
+            else:
+                break  # 遇到盈利就停止计数
+
+        if losing_streak >= 3:
+            return {
+                'losing_streak': losing_streak,
+                'should_cool_down': True,
+                'margin_multiplier': 0.3,
+                'reason': f"连败 {losing_streak} 次，降温（仓位×0.3）"
+            }
+        elif losing_streak >= 2:
+            return {
+                'losing_streak': losing_streak,
+                'should_cool_down': True,
+                'margin_multiplier': 0.5,
+                'reason': f"连败 {losing_streak} 次，降温（仓位×0.5）"
+            }
+
+        return {'losing_streak': losing_streak, 'should_cool_down': False, 'margin_multiplier': 1.0, 'reason': ''}
+
     def _check_risk_control(self, signal: Dict[str, Any],
                            platform_name: str,
                            account: Dict[str, Any],
@@ -2456,6 +2549,22 @@ class CryptoAgent:
             except:
                 pass  # 价格解析失败，跳过检查
 
+        # 5. 资金费率检查（极端资金费率时拒绝开仓）
+        funding_data = signal.get('funding_rate_data')
+        if funding_data:
+            fr_pct = funding_data.get('funding_rate_percent', 0) or 0
+            signal_action = signal.get('action', '')
+            if signal_action == 'buy' and fr_pct > 0.05:
+                return False, f"资金费率过热 {fr_pct:.4f}%（做多拥挤），拒绝做多"
+            elif signal_action == 'sell' and fr_pct < -0.05:
+                return False, f"资金费率过冷 {fr_pct:.4f}%（做空拥挤），拒绝做空"
+
+        # 6. 连败降温检查（不拒绝，但降低仓位）
+        streak_info = self._check_losing_streak(platform_name)
+        if streak_info.get('should_cool_down'):
+            signal['_streak_margin_multiplier'] = streak_info['margin_multiplier']
+            logger.warning(f"[{platform_name}] ⚠️ {streak_info['reason']}")
+
         return True, "通过风控检查"
 
     def execute_signal_with_rules(self, signal: Dict[str, Any],
@@ -3212,6 +3321,77 @@ class CryptoAgent:
         except Exception as e:
             logger.error(f"[Bitget] 检查挂单 TP/SL 补设异常: {e}")
 
+    async def _check_bitget_missing_tp_sl(self):
+        """定时检查 Bitget 持仓是否缺少止盈止损，缺少则从信号补救"""
+        if not self.bitget:
+            return
+        try:
+            positions = self.bitget.get_open_positions()
+            if not positions:
+                return
+
+            for pos in positions:
+                symbol = pos.get('symbol', '')  # e.g. "BTCUSDT"
+                if not symbol:
+                    continue
+
+                # 获取当前止盈止损
+                coin = symbol.replace('USDT', '')
+                tp_sl = self.bitget.get_tp_sl_prices(coin)
+                has_tp = tp_sl.get('take_profit') is not None
+                has_sl = tp_sl.get('stop_loss') is not None
+
+                if has_tp and has_sl:
+                    continue  # 都有，跳过
+
+                # 缺少 TP 或 SL，从信号数据库查找最近信号补救
+                latest_signal = self.signal_db.get_latest_signal('crypto', symbol)
+                if not latest_signal:
+                    missing = ('止盈' if not has_tp else '') + ('/' if not has_tp and not has_sl else '') + ('止损' if not has_sl else '')
+                    logger.warning(f"[Bitget] ⚠️ {symbol} 缺少{missing}，且无历史信号可补救")
+                    continue
+
+                tp_price = latest_signal.get('take_profit')
+                sl_price = latest_signal.get('stop_loss')
+
+                if not tp_price and not sl_price:
+                    logger.warning(f"[Bitget] ⚠️ {symbol} 缺少止盈止损，最近信号也无 TP/SL")
+                    continue
+
+                # 只补救缺少的
+                set_tp = tp_price if not has_tp else None
+                set_sl = sl_price if not has_sl else None
+
+                missing_parts = []
+                if not has_tp:
+                    missing_parts.append(f"TP={set_tp}")
+                if not has_sl:
+                    missing_parts.append(f"SL={set_sl}")
+                missing_desc = ' & '.join(missing_parts)
+
+                logger.warning(f"[Bitget] 🔧 {symbol} 缺少 {missing_desc}，从信号补救...")
+
+                # 用仓位实际大小设置
+                size = abs(pos.get('size', 0))
+                if size <= 0:
+                    continue
+
+                tp_sl_result = self.bitget.set_tp_sl(
+                    symbol=coin,
+                    is_long=pos.get('size', 0) > 0,
+                    size=size,
+                    tp_price=set_tp,
+                    sl_price=set_sl,
+                )
+
+                if tp_sl_result.get('success'):
+                    logger.info(f"[Bitget] ✅ 补救成功: {symbol} {missing_desc}")
+                else:
+                    logger.warning(f"[Bitget] ⚠️ 补救失败: {tp_sl_result.get('error')}")
+
+        except Exception as e:
+            logger.error(f"[Bitget] 止盈止损兜底检查异常: {e}")
+
     def _calculate_hyperliquid_position_size(self, decision: Dict[str, Any], current_price: float) -> float:
         """
         计算 Hyperliquid 仓位大小（基于可用保证金和风控限制）
@@ -3858,12 +4038,19 @@ class CryptoAgent:
     async def _check_position_management_all_platforms(self):
         """检查各平台的持仓管理（止盈/止损/移动止损）"""
         try:
-            # 获取当前价格
+            # 获取当前价格和 ATR 波动率
             current_prices = {}
+            volatility_data = {}
             for symbol in self.symbols:
                 try:
                     data = self.exchange.get_multi_timeframe_data(symbol)
                     current_prices[symbol] = float(data['5m'].iloc[-1]['close'])
+                    # 提取 1h ATR 占价格的百分比，用于动态移动止损
+                    if '1h' in data and 'atr' in data['1h'].columns:
+                        atr_value = data['1h']['atr'].iloc[-1]
+                        price_1h = data['1h']['close'].iloc[-1]
+                        if atr_value and price_1h > 0:
+                            volatility_data[symbol] = float(atr_value) / float(price_1h)
                 except:
                     continue
 
@@ -3881,8 +4068,8 @@ class CryptoAgent:
                 if not positions:
                     continue
 
-                # 检查持仓管理
-                actions = executor.check_position_management(positions, current_prices)
+                # 检查持仓管理（传递 ATR 波动率数据）
+                actions = executor.check_position_management(positions, current_prices, volatility_data)
 
                 # 执行建议的操作
                 for action_info in actions:

backend/app/crypto_agent/executor/base_executor.py

@@ -174,10 +174,16 @@ class BaseExecutor(ABC):
 
     def check_position_management(self,
                                   positions: List[Dict],
-                                  current_prices: Dict[str, float]) -> List[Dict[str, Any]]:
+                                  current_prices: Dict[str, float],
+                                  volatility_data: Optional[Dict[str, float]] = None) -> List[Dict[str, Any]]:
         """
         持仓管理检查
 
+        Args:
+            positions: 持仓列表
+            current_prices: 当前价格 {symbol: price}
+            volatility_data: 波动率数据 {symbol: atr_pct}（1h ATR / price）
+
         Returns:
             建议的操作列表
         """
@@ -230,23 +236,56 @@ class BaseExecutor(ABC):
                     'priority': 2
                 })
 
-            # 规则3: 移动止损
-            if pnl_pct >= 2:
+            # 规则3: 动态移动止损（基于 ATR 波动率）
             current_sl = pos.get('stop_loss')
-                if side == 'buy' and current_sl and current_sl < entry_price:
+            if not current_sl:
+                continue
+
+            # 获取波动率：ATR 占价格的百分比（如 0.015 = 1.5%）
+            atr_pct = (volatility_data.get(symbol) if volatility_data else None) or 0.02
+            # 波动率倍数：低波动（震荡）给更多空间，高波动（趋势）收得更紧
+            atr_multiplier = max(1.5, 2.0 / atr_pct)
+
+            # 两级移动止损
+            # 第一级：盈利 >= 0.6 * ATR% * multiplier → SL 移到入场价 ± 1倍ATR（给足震荡空间）
+            # 第二级：盈利 >= 1.0 * ATR% * multiplier → SL 移到保本
+            step1_threshold = atr_pct * 100 * atr_multiplier * 0.6
+            step2_threshold = atr_pct * 100 * atr_multiplier * 1.0
+
+            if pnl_pct >= step2_threshold:
+                # 盈利充足，移到保本
+                if (side == 'buy' and current_sl < entry_price) or \
+                   (side == 'sell' and current_sl > entry_price):
                     actions.append({
                         'symbol': symbol,
                         'action': 'MOVE_SL',
                         'new_sl': entry_price,
-                        'reason': f"盈利 {pnl_pct:.1f}% >= 2%，移动止损到入场价",
+                        'pnl_pct': pnl_pct,
+                        'reason': f"盈利 {pnl_pct:.1f}% >= {step2_threshold:.1f}%，移动止损到保本",
                         'priority': 3
                     })
-                elif side == 'sell' and current_sl and current_sl > entry_price:
+            elif pnl_pct >= step1_threshold:
+                # 盈利初步，移到入场价 ± 1倍ATR（保留震荡空间，不急于保本）
+                if side == 'buy' and current_sl < entry_price:
+                    new_sl = entry_price * (1 - atr_pct)
+                    if new_sl > current_sl:
                         actions.append({
                             'symbol': symbol,
                             'action': 'MOVE_SL',
-                        'new_sl': entry_price,
-                        'reason': f"盈利 {pnl_pct:.1f}% >= 2%，移动止损到入场价",
+                            'new_sl': new_sl,
+                            'pnl_pct': pnl_pct,
+                            'reason': f"盈利 {pnl_pct:.1f}% >= {step1_threshold:.1f}%，移动止损到入场价-1ATR (${new_sl:.2f})",
+                            'priority': 3
+                        })
+                elif side == 'sell' and current_sl > entry_price:
+                    new_sl = entry_price * (1 + atr_pct)
+                    if new_sl < current_sl:
+                        actions.append({
+                            'symbol': symbol,
+                            'action': 'MOVE_SL',
+                            'new_sl': new_sl,
+                            'pnl_pct': pnl_pct,
+                            'reason': f"盈利 {pnl_pct:.1f}% >= {step1_threshold:.1f}%，移动止损到入场价+1ATR (${new_sl:.2f})",
                             'priority': 3
                         })
 

backend/app/crypto_agent/market_signal_analyzer.py

@@ -208,7 +208,7 @@ class MarketSignalAnalyzer:
             news_context = await self._get_news_context(symbol)
 
             # 3. 获取合约市场数据（资金费率、持仓量等）
-            futures_context = await self._get_futures_context(symbol)
+            futures_context, futures_market_data = await self._get_futures_context(symbol)
 
             # 4. 将日内和趋势拆成两次独立分析，避免一个 prompt 同时混做两件事
             intraday_prompt = self._build_analysis_prompt(
@@ -251,7 +251,24 @@ class MarketSignalAnalyzer:
             intraday_result = self._parse_llm_response(intraday_response or "", symbol)
             trend_result = self._parse_llm_response(trend_response or "", symbol)
 
-            return self._merge_lane_results(symbol, intraday_result, trend_result)
+            result = self._merge_lane_results(symbol, intraday_result, trend_result)
+
+            # 携带量化 regime 数据到最终结果，供执行层使用
+            if market_context.get('range_metrics'):
+                result['range_metrics'] = market_context['range_metrics']
+
+            # 携带资金费率数据到最终结果，供执行层风控使用
+            if futures_market_data:
+                funding = futures_market_data.get('funding_rate') or {}
+                result['funding_rate_data'] = {
+                    'funding_rate': funding.get('funding_rate'),
+                    'funding_rate_percent': funding.get('funding_rate_percent', 0),
+                    'sentiment': funding.get('sentiment'),
+                    'sentiment_level': funding.get('sentiment_level'),
+                    'open_interest': futures_market_data.get('open_interest'),
+                }
+
+            return result
 
         except Exception as e:
             logger.error(f"市场信号分析失败: {e}")
@@ -385,6 +402,7 @@ class MarketSignalAnalyzer:
             'trend': "\n".join(trend_parts),
             'levels': "\n".join(levels_parts),
             'range_warning': range_warning,
+            'range_metrics': range_metrics,
         }
 
     def _get_session_open(self, df: Optional[pd.DataFrame]) -> Optional[float]:
@@ -1184,8 +1202,12 @@ class MarketSignalAnalyzer:
             logger.warning(f"获取新闻失败: {e}")
             return "新闻获取失败"
 
-    async def _get_futures_context(self, symbol: str) -> str:
-        """获取合约市场数据（资金费率、持仓量、溢价率）"""
+    async def _get_futures_context(self, symbol: str) -> tuple:
+        """获取合约市场数据（资金费率、持仓量、溢价率）
+
+        Returns:
+            (formatted_str, raw_market_data) - 格式化文本和原始数据
+        """
         try:
             loop = asyncio.get_event_loop()
             market_data = await loop.run_in_executor(
@@ -1194,11 +1216,11 @@ class MarketSignalAnalyzer:
                 symbol
             )
             if not market_data:
-                return ""
-            return self._format_futures_context(symbol, market_data)
+                return "", None
+            return self._format_futures_context(symbol, market_data), market_data
         except Exception as e:
             logger.warning(f"获取 {symbol} 合约数据失败: {e}")
-            return ""
+            return "", None
 
     def _format_futures_context(self, symbol: str, market_data: Dict[str, Any]) -> str:
         """格式化高价值合约特征，避免大段说明性文本"""