基于小波变换的高压开关柜监测信号消噪方法

为消除智能化高压开关柜监测装置采集信号中的“毛刺”,并克服传统低通滤波方法去除“毛刺”的同时也去掉了有用的高频成分的问题 ,该文提出了一种具有时域选择性的消除“毛刺”的方法。该方法通过利用小波变换具有时频局部化的特点 ,可以准确地检测“毛刺”出现的位置并加以消除 ,从而在消除“毛刺”的同时保留了有用的高频成分。通过对线圈电流的处理结果 ,可看出小波变换对传统低通滤波方法的优势。利用小波检测信号突变点的优势 ,还设计了检测信号起止点的算法

Swichgear signal noise elimination based on wavelet transformation

A new method based on wavelet transform was developed to eliminate the noise among signals which are sampled by the detection device for power switchgear without the disadvantages of conventional low pass filters that indiscriminatingly filter the desirable high frequencies. The new method selectively filters noise. The new method locates noise and eliminates the noise while preserving the valuable high frequencies. In addition, an alogrithm is introduced to detect the beginning and end of signals using wavelet sigularity detection.