小波去噪方法在开关柜故障检测中的应用

Application of Wavelet De-Noising in Fault Detection of MV Switchgear

中压开关柜故障检测中,传感器的输出信号中往往包含大量的噪声干扰,特别是在故障发展的早期阶段。小波变换是常用的降噪处理方法,在高压设备局部放电检测中有很长时间的研究历史。首先介绍了小波变换原理,并对其与傅里叶变换的区别进行了分析,同时介绍了常用的小波降噪的实现方法;其次,搭建了一套开关柜故障测试平台,采用一种简单、实用的微分电场传感器实现了对绝缘子沿面电晕放电信号的测量,并采用一种商用高频CT进行对比测量;最后,基于文中所述小波去噪的实现方法,对两种传感器的输出信号进行了小波去噪处理,原始信号中80 MHz以上频带的干扰信号在小波去噪后得到了有效抑制,小波去噪的效果显著。

In fault detection system for MV switchgear, sensor output signal often contains a lot of noise interference, especially in early stages of fault development. Wavelet transform is a commonly used method for noise reduction. It has a long history of research on partial discharge detection of HV equipment. In this paper, principle of wavelet transform is presented, and its difference with Fourier transform is simply analyzed. Realization method of wavelet de-noising is also presented. A fault test platform of switchgear is built. Simple practical differential electric field sensor is used to measure corona discharge signal along insulator surface, and a commercial high frequency CT is used for comparison. Finally, based on the wavelet de-noising method described in this paper, output signals of two kinds of sensors are de-noised with wavelet, and interference signal above 30 MHz within original signal is effectively suppressed with wavelet de-noising. Effect of wavelet de-noising is very significant.