应用形态学滤波方法去除高压断路器线圈电流信号的噪声

Noise Removal from Coil Current Signal of High Voltage Circuit Breaker with the Morphological Filter

高压断路器是电力系统中重要的控制和保护元件,其可靠性对电力系统的安全具有十分重要的意义。高压断路器机械状态监测能够在一定程度上识别断路器的机械状态。合闸线圈电流信号就是机械状态监测信号之一,合闸线圈电流信号包含有重要的时间和状态信息。笔者在数学形态学理论中的基本形态滤波器基础之上构造形态级联滤波算子对线圈电流信号进行滤波去噪,成功地剔除掉了线圈电流信号中存在的尖峰和沟壑,并且最大限度地保留了电流信号波形的结构特征,使滤波后的电流信号波形平滑性较好、误差较小,便于后续的信号特征提取和状态识别。文中在MATLAB平台中针对上述算法编写了相应的程序,实现了对线圈电流信号的滤波。

High voltage circuit breaker is an important utility in power system, which can protect and control the power lines, so its reliability is vital to the safety of the power system. Mechanical condition monitoring system can identify the mechanical conditions of circuit breakers to some extent. Current signal of the closing coil contains important time and condition information, thus it is applicable for the condition monitoring. In this paper, a morphological filter based on the theory of mathematical morphology is proposed to denoise the coil current signal. This filter can completely eliminate the peaks and valleys of coil current waveform, and keep the characteristic structure of the current waveform as well. The coil current waveform after filtering is very smooth without sacrificing the accuracy and information content of the signal. Corresponding program is coded on the platform MATLAB to implement the proposed filtering algorithm.