基于改进小波变换的电气一次设备故障检测研究

Modified Wavelet Transform-based Fault Detection of Electrical Primary Equipment

传统的电气一次设备故障检测方法通常基于固定的阈值进行故障判断,导致故障检测的误报率较高。针对上述问题,提出基于改进小波变换的电气一次设备故障检测方法。首先,通过监测电气一次设备的状态,实时获取设备的运行数据。接下来,利用改进小波变换对故障信号进行分解,将复杂的信号分解为多个尺度的成分。最后,根据信号分解与特征熵值从大量的数据中自动提取出与故障相关的特征,利用这些特征进行故障检测。实验结果表明,基于改进小波变换的电气一次设备故障检测方法的误报率低,检测结果具有较高的可信度,具有应用价值。

Conventional electrical primary equipment fault detection methods generally use fixed thresholds for fault diagnosis,and the false alarm rate is relatively high.To address this problem,a fault detection method based on modified wavelet transform is proposed.First,by monitoring the status of electrical primary equipment,real-time operational data of the equipment can be obtained.Then by using the modified wavelet transform,the complex fault signal is decomposed into components on multiple scales.By utilizing signal decomposition and feature entropy values,automatic extraction of fault-related features for fault detection from the massive data is ultimately achieved.The proposed method has been verified through experiments to achieve low false alarm rate and reliable detection result for electrical primary equipment faults.