diff --git a/stock-html/routes/analysis.py b/stock-html/routes/analysis.py
index bf504c5..325b334 100644
--- a/stock-html/routes/analysis.py
+++ b/stock-html/routes/analysis.py
@@ -99,7 +99,7 @@ def deep_analyze():
         if not stock_code:
             return jsonify({'error': '股票代码不能为空'}), 400
 
-        df = algo_get_kline_data(stock_code, days=180)
+        df = algo_get_kline_data(stock_code, days=120)
         if df is None or len(df) < 30:
             return jsonify({'error': 'K线数据不足'}), 400
 
diff --git a/stock-html/services/score_engine.py b/stock-html/services/score_engine.py
index ca0112a..8fa489c 100644
--- a/stock-html/services/score_engine.py
+++ b/stock-html/services/score_engine.py
@@ -134,10 +134,16 @@ def compute_comprehensive_score(stock_code, stock_name, technical_score, df=None
     external_score = 0
     summaries = []
 
-    # ---- P0: 主力资金进出 ----
+    # ---- P0: 主力资金进出（带30分钟缓存，与批量模式一致）----
     try:
-        from services.fund_flow_analyzer import analyze_fund_flow
-        fund_result = analyze_fund_flow(stock_code, days=5)
+        now = time.time()
+        cached_ff = _fund_flow_cache.get(stock_code)
+        if cached_ff and (now - cached_ff[1]) < _FUND_FLOW_TTL:
+            fund_result = cached_ff[0]
+        else:
+            from services.fund_flow_analyzer import analyze_fund_flow
+            fund_result = analyze_fund_flow(stock_code, days=5)
+            _fund_flow_cache[stock_code] = (fund_result, now)
         factors['fund_flow'] = fund_result
         external_score += fund_result.get('score', 0)
         all_reasons.extend(fund_result.get('reasons', []))
diff --git a/stock-html/services/signal_detector.py b/stock-html/services/signal_detector.py
index c31947b..38d26a2 100644
--- a/stock-html/services/signal_detector.py
+++ b/stock-html/services/signal_detector.py
@@ -340,7 +340,9 @@ def detect_all_signals(df, lookback=5):
         if not np.issubdtype(df[col].dtype, np.floating):
             df[col] = pd.to_numeric(df[col], errors='coerce').fillna(0).astype(np.float64)
 
-    df = calc_all_indicators(df)
+    # 若已有指标列则跳过重复计算（deep_analyze 等场景预计算过）
+    if 'dif' not in df.columns or 'ma5' not in df.columns:
+        df = calc_all_indicators(df)
 
     all_signals = []
     all_signals.extend(detect_main_rising_wave(df, lookback))
diff --git a/stock-html/services/stock_algorithms.py b/stock-html/services/stock_algorithms.py
index 939fb97..0ff2e48 100644
--- a/stock-html/services/stock_algorithms.py
+++ b/stock-html/services/stock_algorithms.py
@@ -172,7 +172,6 @@ def _get_kline_from_local_db(stock_code, days=120):
         conn = get_db()
         if not conn:
             return None
-        conn.autocommit = True
         start_date = (datetime.now() - timedelta(days=days)).strftime('%Y-%m-%d')
         try:
             with conn.cursor() as cur:
diff --git a/stock-html/services/stock_service.py b/stock-html/services/stock_service.py
index 8bd309a..19d8fa1 100644
--- a/stock-html/services/stock_service.py
+++ b/stock-html/services/stock_service.py
@@ -7,11 +7,16 @@ from datetime import datetime, timedelta
 import traceback
 import json
 import os
+import time
+import threading
 from config import Config
 
 
 # ========== 股票名称缓存 ==========
 _stock_name_cache = {}
+_name_cache_dirty = False
+_name_cache_last_save = 0.0
+_name_cache_lock = threading.Lock()
 
 
 def _load_stock_name_cache():
@@ -26,23 +31,51 @@ def _load_stock_name_cache():
         print(f"加载股票名称缓存失败: {e}")
 
 
-def _save_stock_name_cache():
-    """保存股票名称缓存到本地"""
-    try:
-        with open(Config.STOCK_NAME_CACHE_FILE, 'w', encoding='utf-8') as f:
-            json.dump(_stock_name_cache, f, ensure_ascii=False, indent=2)
-    except Exception as e:
-        print(f"保存股票名称缓存失败: {e}")
+def _save_stock_name_cache(force=False):
+    """保存股票名称缓存到本地（延迟批量保存，60秒去抖）"""
+    global _name_cache_dirty, _name_cache_last_save
+    if not force and not _name_cache_dirty:
+        return
+    now = time.time()
+    if not force and (now - _name_cache_last_save) < 60:
+        return
+    with _name_cache_lock:
+        try:
+            with open(Config.STOCK_NAME_CACHE_FILE, 'w', encoding='utf-8') as f:
+                json.dump(_stock_name_cache, f, ensure_ascii=False, indent=2)
+            _name_cache_dirty = False
+            _name_cache_last_save = now
+        except Exception as e:
+            print(f"保存股票名称缓存失败: {e}")
 
 
 def get_stock_name(stock_code):
-    """获取股票名称 — 使用腾讯财经API"""
-    global _stock_name_cache
-    
+    """获取股票名称 — 优先内存缓存 → DB stock_realtime_price → 腾讯财经API"""
+    global _stock_name_cache, _name_cache_dirty
+
     if stock_code in _stock_name_cache:
         return _stock_name_cache[stock_code]
-    
-    # 腾讯财经API获取股票名称
+
+    # 优先从数据库 stock_realtime_price 表查（毫秒级）
+    try:
+        from db import get_db, put_db
+        conn = get_db()
+        if conn:
+            try:
+                with conn.cursor() as cur:
+                    cur.execute("SELECT name FROM stock_realtime_price WHERE code = %s", (stock_code,))
+                    row = cur.fetchone()
+                    if row and row[0]:
+                        _stock_name_cache[stock_code] = row[0]
+                        _name_cache_dirty = True
+                        _save_stock_name_cache()
+                        return row[0]
+            finally:
+                put_db(conn)
+    except Exception:
+        pass
+
+    # 回退到腾讯财经API
     try:
         import requests as _req
         tcode = ('sh' if stock_code.startswith('6') else 'bj' if stock_code.startswith(('8', '9')) else 'sz') + stock_code
@@ -52,11 +85,12 @@ def get_stock_name(stock_code):
             _fields = _r.text.split('\"')[1].split('~')
             if len(_fields) > 2 and _fields[1]:
                 _stock_name_cache[stock_code] = _fields[1]
+                _name_cache_dirty = True
                 _save_stock_name_cache()
                 return _fields[1]
     except Exception as e:
         print(f"获取股票名称失败(腾讯): {e}")
-    
+
     return None