张红,雷志国,金月,谭万禹.基于SVM的电能质量扰动分类[J].电测与仪表,2014,51(23):. zhanghong,leizhiguo,jinyue,Tanwanyu.Research of Power Quality Disturbance classification Based On SVM[J].Electrical Measurement & Instrumentation,2014,51(23):.
基于SVM的电能质量扰动分类
Research of Power Quality Disturbance classification Based On SVM
There are many electrical power quality signals in power system, It is difficult to extract disturbance signal data. This paper adopts a method of discrete wavelet transformation to extract the power quality disturbance signals; This paper extracts the subtraction of energy on each of the frequency bands as the signal feature vector. decreasing classification of disturbance data.The software of MATLAB is used to generate the disturbance signal samples and the signal-noise ratio of 25dB (SNR=25 dB) Gauss white noise was added in the samples. The libsvm is used to solve multi-class SVM classification problems and a two-step grid search method is presented to optimize SVM parameters. The simulation results show the correct identification and classification of the presented method has a higher rate and it also verifies the correctness and effectiveness of this method.