Adaptive Suppression Method for Periodic Pulsation Interference in Partial Discharge of Converter Transformers Based on Periodic Consistency Scoring and Waveform Characteristics

The online monitoring of partial discharge (PD) in converter transformers faces significant noise interference. Among these, the periodic pulsating interference caused by the switching process of thyristors is particularly challenging to directly identify from waveforms or frequency spectra, which affects the accuracy of PD monitoring. To address this, a time-domain multi-feature joint recognition algorithm based on periodic consistency scoring and waveform characteristics is proposed. Firstly, the timestamps of each pulse in the signal sequence are extracted based on the cumulative energy function. Secondly, the periodic consistency score and time–frequency feature parameters of each pulse are extracted separately, and an optimal recognition vector is constructed based on univariate screening and redundancy feature checking. Finally, dimensionality reduction is performed using principal component analysis, and the PD signals are separated from periodic pulse interference through fuzzy C-means clustering. The proposed algorithm is applied to denoise PD signals superimposed with on-site noise interference and field-measured signals. The results demonstrate that the algorithm effectively suppresses periodic pulsation interference while significantly reducing the attenuation of PD signals during the denoising process.