Self-Adaptive Wavelet Based on Parametric Equation in Manufacturing Engineering

Abstract

The paper analyzes the difference between useful signal and noise signal in dissemination characteristic inside wavelet space in manufacturing engineering, and then puts forward a search algorithm based on wavelet decorrelation white noise testing and involved with the parameters in a parametric equation. The algorithm can select wavelet transform to realize the best noise reduction effect in a self-adaptive way according to the characteristic of signal containing noise and signal to noise ratio. At last, simulation experiment and engineering application are made, and their results are compared with the decomposition result of Daubechies wavelet. It’s concluded that self-adaptive wavelet basis can more adequately separate useful information from signal.