From 39cdaa4d7519e62cfb94cf9833aa8ec40e47373a Mon Sep 17 00:00:00 2001
From: aaron <>
Date: Sat, 21 Feb 2026 19:36:58 +0800
Subject: [PATCH] =?UTF-8?q?=E4=BF=AE=E5=A4=8D=E7=8E=B0=E6=88=90=E9=98=BB?=
 =?UTF-8?q?=E5=A1=9E=E7=9A=84=E9=97=AE=E9=A2=98=E3=80=82?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .../app/crypto_agent/llm_signal_analyzer.py   |  8 ++---
 backend/app/crypto_agent/signal_analyzer.py   |  4 +--
 backend/app/services/llm_service.py           | 36 ++++++++++++++++++-
 3 files changed, 41 insertions(+), 7 deletions(-)

diff --git a/backend/app/crypto_agent/llm_signal_analyzer.py b/backend/app/crypto_agent/llm_signal_analyzer.py
index b28db86..5925031 100644
--- a/backend/app/crypto_agent/llm_signal_analyzer.py
+++ b/backend/app/crypto_agent/llm_signal_analyzer.py
@@ -578,8 +578,8 @@ class LLMSignalAnalyzer:
             # 构建数据提示
             data_prompt = self._build_data_prompt(symbol, data, news_text, position_info, futures_data)
 
-            # 调用 LLM
-            response = llm_service.chat([
+            # 调用 LLM（使用异步方法避免阻塞事件循环）
+            response = await llm_service.achat([
                 {"role": "system", "content": system_prompt},
                 {"role": "user", "content": data_prompt}
             ], model_override=self.model_override)
@@ -1719,8 +1719,8 @@ class LLMSignalAnalyzer:
             # 构建持仓分析提示
             prompt = self._build_position_review_prompt(symbol, positions, data)
 
-            # 调用 LLM
-            response = llm_service.chat([
+            # 调用 LLM（使用异步方法避免阻塞事件循环）
+            response = await llm_service.achat([
                 {"role": "system", "content": self.POSITION_REVIEW_PROMPT},
                 {"role": "user", "content": prompt}
             ], model_override=self.model_override)
diff --git a/backend/app/crypto_agent/signal_analyzer.py b/backend/app/crypto_agent/signal_analyzer.py
index a97ecfa..47ef1d3 100644
--- a/backend/app/crypto_agent/signal_analyzer.py
+++ b/backend/app/crypto_agent/signal_analyzer.py
@@ -1089,8 +1089,8 @@ class SignalAnalyzer:
             # 构建分析提示
             prompt = self._build_analysis_prompt(data, signal, symbol)
 
-            # 调用 LLM
-            response = llm_service.chat([
+            # 调用 LLM（使用异步方法避免阻塞事件循环）
+            response = await llm_service.achat([
                 {"role": "system", "content": self.CRYPTO_ANALYST_PROMPT},
                 {"role": "user", "content": prompt}
             ])
diff --git a/backend/app/services/llm_service.py b/backend/app/services/llm_service.py
index c05bdba..45e6914 100644
--- a/backend/app/services/llm_service.py
+++ b/backend/app/services/llm_service.py
@@ -1,6 +1,7 @@
 """
 LLM服务 - 兼容层，使用多模型服务
 """
+import asyncio
 from typing import Optional, List, Dict, Any
 from app.services.multi_llm_service import multi_llm_service
 from app.utils.logger import logger
@@ -23,7 +24,7 @@ class LLMService:
         model_override: str = None
     ) -> Optional[str]:
         """
-        调用LLM进行对话
+        调用LLM进行对话（同步版本）
 
         Args:
             messages: 消息列表
@@ -42,6 +43,39 @@ class LLMService:
             model_override=model_override
         )
 
+    async def achat(
+        self,
+        messages: List[Dict[str, str]],
+        model: str = None,
+        temperature: float = 0.7,
+        max_tokens: int = 2000,
+        model_override: str = None
+    ) -> Optional[str]:
+        """
+        调用LLM进行对话（异步版本 - 不阻塞事件循环）
+
+        Args:
+            messages: 消息列表
+            model: 模型名称（已废弃，使用 model_override）
+            temperature: 温度参数
+            max_tokens: 最大token数
+            model_override: 指定使用的模型 (zhipu/deepseek)
+
+        Returns:
+            LLM响应文本
+        """
+        # 在线程池中执行同步的 LLM 调用，避免阻塞事件循环
+        loop = asyncio.get_event_loop()
+        return await loop.run_in_executor(
+            None,
+            lambda: self.multi_service.chat(
+                messages=messages,
+                temperature=temperature,
+                max_tokens=max_tokens,
+                model_override=model_override
+            )
+        )
+
     def analyze_intent(self, user_message: str) -> Dict[str, Any]:
         """使用LLM分析用户意图"""
         return self.multi_service.analyze_intent(user_message)