Recognition of transient power quality disturbances based on CWD spectral kurtosis

Since the process of present transient power quality disturbance recognition is very complex and sensitive to noise,an algorithm based on CWD spectral kurtosis and combined with RMS is proposed for its recognition,which sorts the disturbance signals according to the CWD spectral kurtosis into three types: transient pulse,transient oscillation and amplitude disturbance,and then classifies the amplitude disturbance according to its RMS into one of three voltage kinds:swell,sag and interrupt. It does not need any classifier,extremely simplifying the calculation process and reducing the calculation time. Simulative results verify the correctness of theoretical analysis.