The application of power quality signal de-noising based on the improved wavelet threshold function

Abstract

The contradiction between de-noising and preserving the local features of the original signal simultaneously cannot be solved with the traditional wavelet threshold function (TF), and thus an improved combination of soft and hard threshold functions (TFs) of power quality signal de-noising is proposed in this paper. The improved wavelet TF is utilized to overcome the problem of “Pseudo-Gibbs phenomenon”, the deviation of wavelet coefficients, caused by the traditional wavelet TF, which fills a gap for the deficiency of wavelet soft and hard threshold de-noising, and with stronger adaptability. Finally, the de-noising tests of three typical signals of power quality are simulated based on MATLAB R2016a. Compared with the traditional wavelet TFs, the results obtained based on the proposed improved wavelet TF can guarantee better effectiveness and feasibility.