Optimum wavelet bases selection for wavelet based de-noising in partial discharge measurement

In this paper a new approach based on wavelet transform to select optimum wavelet bases for on line Partial Discharge (PD) measurement is proposed. Among the previous work correlation coefficient and signal-to-noise ratio is used for evaluating the optimal wavelet bases. In statistical point of view, these two indexes can not give optimum de-noising performance comprehensively. So, in this paper, the concept of energy loss is taken into the account and it is more effective for further analyzing the quality of partial discharge signal. This method is applied on two simulated PD signal and also on field detected signal. From the results it shows that the de-noising procedure can be improved by selecting the optimal wavelet bases.