Advanced processing techniques of high-voltage impulse test signals

Automated analysis and processing of the digital records obtained in high-voltage (HV) impulse tests requires special techniques that are able to detect and characterize all sorts of disturbances (of short and long duration) in all waveform types (full, chopped, and sliced). Especially for the cases where a short-duration disturbance has to be removed before the evaluation of the impulse parameters, it is important to characterize the disturbance both in time and frequency. This paper describes a new technique based on the wavelet transform used for this purpose. Experimental results show the technique's reliability when applied to actual HV impulses.