自适应提升小波变换及飞机发动机故障分析

Adaptive lifting wavelet transform and aeroengine fault analysis

小波提升方案通过改变预测器和更新器构造出所需要的小波,这为机械设备故障特征分析中小波基函数的选择提供了方便。为了能在每个尺度上自适应选择与机械振动信号特征匹配的小波基函数,提出了一种更新器和预测器同时自适应地提升小波变换方法。在此方法中,采用先更新后预测的提升方案,分别通过信号的局部梯度大小和最小化预测误差来实现自适应更新和预测。将此方法应用在某飞机发动机故障分析中,实验结果表明,与经典小波变换相比该自适应提升小波变换分离的故障特征更明显效果更好。

The lifting scheme can only change the predictor and updater to construct the appropriate wavelet,which facilitates the wavelet function selecting in the mechanical equipment fault analysis.In order to adaptively select the wavelet function matching the signal feature better,an update/prediction adaptive lifting wavelet transform is proposed.In the method,an adaptive update/ prediction lifting scheme is used,which realizes adaptive update and prediction respectively by the local gradient information of original signal and minimum prediction error.The experimental results show the proposed method plays more effectively in failure analysis of aeroengine signal than classical wavelet transform.