Harmonic and interharmonic detection based on adaptive wavelet and improved EWT

Abstract

To address the issues of noise sensitivity and over-decomposition when decomposing harmonic signals using empirical wavelet transform （EWT）， a method based on adaptive wavelet denoising and improved EWT is proposed for detecting harmonics and interharmonics. Firstly， to enhance the effectiveness of traditional wavelet threshold functions， an adaptive parameterless threshold function is introduced. Then， by analyzing the signal spectrum using EWT， a series of filter banks is obtained to decompose the signal components. Finally， dynamic time warping （DTW） is employed to reconstruct the over-decomposed signals， resulting in the final components and identifying their frequencies and amplitudes. Simulation results demonstrate that the proposed method effectively suppresses noise in the harmonics and improves the over-decomposition of EWT. Comparison with empirical mode decomposition （EMD） and particle swarm optimization based variational mode decomposition （PSO-VMD） verifies the superiority of the proposed method in separating harmonics and detecting interharmonics.

Keywords

• harmonic
• interharmonic
• ewt
• wavelet threshold function
• dtw