Application of FWT (fast wavelet transform) for auto-detection system of partial discharge in power cables

Conclusively, the statistic results of the data of partial discharges and noise signals transformed by employing FWT (fast wavelet transform) revealed the characteristics of the waveforms in time-frequency space. The authors effectively succeeded to obtain those data that show them the distinguished properties. The significance of this analysis is that they can observe not only the frequency distribution of the spectrum, but also the time distribution of it at the same time. They expect that it will lead a neural network (NN) system to executing the identification or partial discharges in underground power CV cables. The results in employing the NN system with FWT show some aspects that the FWT will be the best method for waveform pattern recognition.