改进的小波变换在罗拉故障诊断中的应用

Application of Improved Wavelet Transform in the Roller Fault Diagnosis

针对纺织细纱机中罗拉故障的诊断问题,采用了一种改进的小波分解与重构算法对故障进行特征提取,并通过反复的实验选取出了比较合适的双正交小波基以准确地提取特征信息;分析了传统小波变换在罗拉故障中产生混频的原因,给出了一种改进的小波算法。该算法将小波变换与FFT相结合以消除混频现象。仿真研究表明,双正交小波基对于提取罗拉故障的特征频率有较好的效果,改进的小波算法能够较好地消除频率混叠现象。

Aiming at the roller fault diagnosis in textile spinning frames,the improved method that adopting wavelet analysis and reconstruction algorithm is used to extract the features of faults.In order to extract more precise feature information,through repeating experiments,more suitable bi-orthogonal wavelet basis is selected.The causes of frequency overlapping found in application of traditional wavelet transform for roller fault are analyzed,and the improved wavelet algorithm is proposed.This algorithm combines wavelet transform and FFT to eliminate the phenomenon of frequency overlapping.The simulation research shows that the bi-orthogonal wavelet basis is more effective for extracting feature frequency of roller fault;and the improved wavelet algorithm well eliminates the frequency overlapping.