Denoising of UHF PD signals based on optimised VMD and wavelet transform

The ultra-high frequency (UHF) partial discharge (PD) monitoring is one of the most effective ways to detect the insulation failure in electrical equipment. Although the UHF method avoids the intensive low-frequency noise, various electromagnetic interferences in substation can still influence its effectiveness and reliability. Most existing denoising algorithms, however, are unable to suppress different kinds of noise simultaneously. Meanwhile, the main focus of these algorithms is on the degree of noise reduction, while pay little attention to feature preservation. Therefore, a novel denoising method namely optimised variation mode decomposition and wavelet (OVMDW) is developed in this paper. In OVMDW, an optimised variational mode decomposition (VMD) algorithm is presented firstly to decompose the original signal into several band-limited modes. Then, a determination rule is designed to identify the effective components from these modes. Finally, the wavelet-based denoising method is employed to remove the residual white noise. Denoising results for simulative and experimental data show that the proposed method can remove various interferences from UHF PD signals. Additionally, new indexes are introduced to evaluate the performance of denoising algorithms in practical situations. Comparison results based on these indexes reveal that the proposed method using the VMD approach can preserve quite well the features of original signals.