Transformer partial discharge signal adaptive denoising method based on white shark optimization optimized successive...

Area of Science:

• Electrical Engineering
• Signal Processing
• Materials Science

Background:

• Partial discharge detection is critical for electrical insulation integrity.
• White noise and periodic narrowband interference degrade signal quality.
• Existing denoising methods may struggle with complex interference patterns.

Purpose of the Study:

• To develop an effective partial discharge denoising method for improved detection accuracy.
• To address limitations of traditional variational mode decomposition (VMD).
• To enhance signal-to-noise ratio while preserving waveform features.

Main Methods:

• Successive Variational Mode Decomposition (SVMD) optimized with a White Shark Optimizer for the balance parameter.
• Kurtosis criterion for mode selection to eliminate periodic narrowband interference.
• Wavelet threshold denoising for residual white noise removal.
• Signal reconstruction for the final denoised output.

Main Results:

• The proposed SVMD combined with wavelet thresholding significantly outperforms existing methods.
• The White Shark Optimizer effectively determines the optimal balance parameter for SVMD.
• Periodic narrowband interference and white noise were substantially reduced.
• Preservation of key partial discharge waveform characteristics was demonstrated.

Conclusions:

• The novel denoising method offers superior performance in partial discharge analysis.
• This approach enhances the reliability of partial discharge detection systems.
• The technique is robust in retaining essential signal information for accurate diagnostics.