小波变换在切削颠振特征提取中的应用

Application of wavelet transform in extracting characteristics of cutting chatter

本文利用小波变换的多分辨率特性,首先对切削颤振信号进行小波分解,然后通过主成分分析法对特征小波进行重构,再对重构信号进行特征提取,从而有效地解决了课题基于隐马尔可夫模型(HMM)切削颤振预报的特征提取问题。文章还提出了从时域基于功率分析对颤振信号进行特征提取的新方法,为切削颤振的特征提取提供了新的模式样本。

This article effectively solves the characteristic extracting problem of the subject-predlctlon of the cutting chatter based on Hidden Markov Model,Firstly decomposing the cutting chatter signal and using the principal composition analysis method to reconstruct a signal from the feature wavelet, next extracting the characteristic from the reconstructed signal. It proposes a new method of characteristic extracting for chatter signal in time domain based on power analysis and provides a new sampling model to the characteristic extracting of cutting chatter.