基于改进快速MP算法的GIS故障分类及定位分析

GIS fault classification and location analysis based on improved fast MP algorithm

随着电网的大范围普及,气体绝缘组合电气(GIS)出现的故障也越来越多,通过对局部放电(PD)检测实现对GIS的故障分类以及定位,PD信号的提取方式主要是小波变换,只能去除噪音,无法提取具体的数据。GIS的故障分类一般情况下都出现在单个PD的信号源中,在多个PD信号源同时存在的情况下不能进行很好的故障检测,在定位的过程中也容易受到信号的干扰,无法获取准确的信息,针对GIS故障分类以及定位技术上的不足,采用改进的快速波形匹配追踪(MP)算法对GIS的故障进行信号的提取。因此研究一套能够迅速、准确的对GIS设备的运行状态进行检测的系统是目前电力企业的主要任务。

With the popularization of the power grid,more and more faults occur in gas insulated combined electrical(GIS).Through partial discharge(PD) detection,the fault classification and location of GIS are realized.The extraction method of PD signal is mainly wavelet transform,which can only remove noise and can not extract specific data.The fault classification of GIS usually occurs in a single PD signal source.It can’t detect the fault very well when there are many PD signal sources at the same time.It is easy to be disturbed by the signal and can’t get accurate information in the process of positioning.The fast waveform matching pursuit(MP) algorithm is used to extract signals from GIS faults.Therefore,it is the main task of power enterprises to study a set of system which can quickly and accurately detect the operation status of GIS equipment.