Optimal wavelet selection to identify faults during impulse tests

The purpose of conducting lightning impulse tests on a transformer is to check the dielectric strength of the transformer insulation against overvoltages of atmospheric origin. Several signal analysis methods have been proposed to overcome the limitations associated with conventional fault detection techniques. In one such group of methods test waveforms are analyzed to localize the disturbance in both time and frequency. There have been several studies on the application of wavelets to identify faults during impulse testing. Since there are a number of wavelets available, the question arises as to how to choose the most appropriate wavelet for a particular application. In this paper, a method based on statistical analysis is used to select the optimum wavelet to detect faults during impulse testing and preliminary results of the research work are presented.