Self-adaption wavelet packet based on improved threshold algorithm

Wavelet threshold de-noising algorithm is an effective method to remove noise in test signals. Based on the analysis of features presented by useful signals and noise signals in the working process of working bearings, a self-adaption of optimal decomposition level algorithm based on improved threshold wavelet packet to remove related white noise test is proposed. This algorithm can adaptively select the optimal decomposition levels that wavelet transforms according to the features and SNR of noise signals, so as to realize the best de-noising effect. It turns out from the engineering test that this algorithm can fully separate the useful information from the signals.