An approach to the extraction and classification of feature vector for power quality based on wavelet

An approach to the extraction and classiflcation of feature vector for power quality based on wavelet is proposed in this paper.The disturbances include the voltage swells,sags,interruptions,transients and notchs.With the singularity principle of wavelet transform,the abrupt changes in the signal of the dynamic power quality are extracted and located,and the time during signal variation is then deflned.Thus,the disturbances can be classifled by the voltage averaging method.The results of MATLAB simulation show its validity.Finally,the proposed method is realized on a digital signal processor(DSP).