Research On Dynamic Power Quality Disturbance Recognition Based On Wavelet Entropy And SVM

In recent years, power quality disturbances detection and classification has become a problem which attracts the concern in many fields. This paper mainly discusses the various models of dynamic power quality disturbances detection and classification, a novel approach on dynamic power quality disturbance classification based on wavelet singular entropy and support vector machine is presented in this paper. At first the wavelet packet transform is applied to reduce the noise of power quality disturbance signal. Then the wavelet singular entropy is used to extract the feature of disturbance signal. At last the pattern recognition classifier based on the SVM algorithm is used to train and classify the eigenvector. The simulation results show that the proposed method is effective in power quality disturbances classification which has the properties of powerful anti-noise performance, simple model, high recognition rate and high generalized performance.