Elevator abnormal sound detection method and system

The invention relates to an elevator abnormal sound detection method and system. The method comprises the following steps of first, acquiring a video record of an elevator car, and extracting audio data from the video record; second, preprocessing the audio data, and extracting audio features from the audio data; third, inputting the audio data into a detection model, and determining whether an elevator generates abnormal sounds or not at the current moment; and fourth, if yes, determining whether an alarm signal is sent out or not according to the amplitude of the abnormal sounds. The elevator abnormal sound detection method and system have lower requirements on hardware equipment, and with the method and system, most abnormal sounds generated by running elevators can be recognized, and the application range is wide. 本发明涉及一种电梯异响检测方法及系统，方法包括以下步骤：a、采集电梯轿厢的视频录像，并从中提取音频数据；b、对所述音频数据进行预处理，并从中提取音频特征；c、将音频数据输入检测模型，识别电梯在当前时刻是否发生异响；d、若判断电梯在当前时刻发生异响，则进而根据异响的幅值大小决定是否发出告警信号。本发明降低了对硬件设备的要求，并且能够识别大多数电梯在运行过程中产生的异响，应用范围较广。