Joint time-frequency analysis for partial discharge (PD) signal extraction

The joint time-frequency analysis has been recently applied to PD signal extraction. Its advantage is widely considered to be the capability of analyzing signals in time and frequency domains at multiple scales. This paper presents a study on recent advances in the application of joint time-frequency analysis techniques to extract PD signal from noise. The theoretical foundation and implementation of two representative joint time-frequency analysis techniques, namely discrete wavelet transform (DWT) and empirical mode decomposition (EMD) are detailed. Case studies using signals taken from PD measurements are presented.