基于深度学习算法的电梯振动异常检测方法

Detection Method of Elevator Vibration Abnormality Based on Deep Learning Algorithm

针对电梯振动异常检测方法存在振动数据处理效果差导致振动异常检测效率及精度低的问题,本文提出基于深度学习算法的电梯振动异常检测方法。通过对电梯在运行过程中产生的振动数据进行预处理,建立数据边缘网格;再引入深度学习算法分析电梯振动异常时序数据,实现对电梯振动异常的快速检测。由实验对比结果可知,改进的检测方法其检测执行时间仅为8.7 ms,检测精度均在95.00%以上,与基于LSTM-VAE的电梯异常检测方法相比,有效提高了检测精度,降低了执行时间,具有一定的优势。

Aiming at the problem of low efficiency and accuracy of elevator vibration anomaly detection due to the poor effect of vibration data processing,an elevator vibration anomaly detection method based on deep learning algorithm is proposed.By preprocessing the vibration data generated during the operation of the elevator,the data edge grid is established;Then the deep learning algorithm is introduced to analyze the time series data of elevator vibration anomalies,so as to realize the rapid detection of elevator vibration anomalies.According to the experimental comparison results,compared with the elevator anomaly detection method based on LSTM-VAE,the detection execution time of the improved detection method is only 8.7 ms,and the detection accuracy is more than 95.00%,which effectively improves the detection accuracy and reduces the execution time,and has certain advantages.