基于第二代小波变换的混合小波降噪方法

Hybrid Wavelet Denosing Approach Using Second Generation Wavelet Transform

根据第二代小波变换的特点,提出一种自适应小波变换的框架,在此基础上提出一种混合小波降噪方法。预测时从一组预测器中选择一个最佳预测器以满足细节信号的特殊要求,更新时从一组更新器中选择一个最佳更新器以满足逼近信号的特殊要求,在每一尺度上构造了一个具有某种要求的小波。这种方法能够有效地抑制振动信号的高频干扰。

By virtue of the second generation wavelet transform, a framework of adaptive wavelet transform was put forward. Based on the framework, a new hybrid wavelet denosing approach was proposed. An optimal predictor was selected from a group of predictors to fulfill the requirements for the detail signals when making predieton,and an optimal updater was selected from a group of updat- ers to fulfill the requirements to the approximate signals in updating stage, then a wavelet with special property was built on a decomposition level. The methed has ability to suppresse high-- frequency noise in vibration signals efficiently.