电能质量暂态扰动模糊分类分析

Fuzzy Classification Analysis of Transient Disturbance of Power Quality

基于电能质量暂态扰动信号的多变特征,提出一种面向一对多的改进型支持向量机的电能质量暂态扰动信号的分类。利用一种改进的小波变换阈值函数来进行信号的分离,和扰动信号的频带分层,实现对电能扰动低频稳定信号和高频噪声信号的分离;采用Daubechise 4小波函数进行电能质量扰动信号的提取,并利用K-CV方法选定相关的扰动信号参数。实例验证表明:基于采用改进的SVM电能质量暂态扰动分类法的单一暂态扰动指标的平均准确率明显优于传统BP神经网络法,采用改进的SVM法能有效识别电网中存在的电能质量暂态扰动。

Based on the variance characteristics of the power quality transient disturbance signals,an improved support vector machine(SVM)is proposed to classify the disturbance.An improved threshold function of wavelet transform is used to separate these signals,and the frequency band of the disturbed signals is stratified to separate the low-frequency stable signals and high-frequency noise signals.Daubechise 4wavelet function is used to extract power quality disturbance signal,and relative disturbance signal parameters are selected by K-CV method.At last,example validation shows that based on the improved SVM classification the average accuracy of a single transient disturbance index of transient power quality disturbance is superior to the traditional BP neural network,the improved SVM method can effectively identify the grid of transient power quality disturbance.