Refactor speech processing: add type hint for `_process_active_voice_frame` and replace `max()` with `np.max()` for consistency.

src/VoiceDialogue/services/speech/speech_monitor.py

@@ -111,7 +111,7 @@ class SpeechStateMonitor(BaseThread):
         """计算音频帧时长（毫秒）"""
         return librosa.get_duration(y=audio_frame, sr=self.sample_rate) * 1000
 
-    def _process_active_voice_frame(self, audio_frame):
+    def _process_active_voice_frame(self, audio_frame: np.ndarray):
         """
         处理活跃语音帧
         
@@ -121,7 +121,7 @@ class SpeechStateMonitor(BaseThread):
         Returns:
             bool: 是否为有效的活跃语音帧
         """
-        if audio_frame.max() <= self.config.MIN_AUDIO_AMPLITUDE:
+        if np.max(audio_frame) <= self.config.MIN_AUDIO_AMPLITUDE:
             return False
 
         # 重置静音计时