Based on SVM Power Quality Disturbance Classification Algorithm

This paper, by using support vector machine (SVM) identification of power quality disturbance signals are classified. In order to obtain better classification results, we need to make a pretreatment for power quality disturbance data. Because wavelet transform has good local characteristics of the processing ability, so the disturbance signal uses wavelet transform to extract the scale of the energy difference as a feature vector. At the same time the Lib - SVM is used to solve the problem of multi class SVM classification, besides, we put forward two steps grid method for SVM parameters optimization. We use MATLAB software to produce disturbance signal data samples and add SNR = 25 db gaussian white noise. Simulation result shows that the proposed classification method of the correct recognition rate is higher, so the correctness and effectiveness of the presented approach are correct and effective.