自适应冗余提升多小波包在滚动轴承复合故障诊断中的应用

A Multi-Fault Diagnosis of Rolling Element Bearing Based on Adaptive Redundant Lifting Multiwavelet Packet Transform

针对多小波具有多个小波函数和尺度函数的特性,在滚动轴承复合故障诊断过程中可以很好地匹配多个故障特征,具备了一次性提取复合故障特征的能力。提出了基于提升框架的多小波包算法。以信号残差最小值为优化目标,对提升小波的预测算子和更新算子进行优化,使得提升小波基能够自适应与观测信号相匹配。通过对输出信号进行后处理,能够将多个故障呈现在不同的通道中,实现复合故障的一次性提取。对滚动轴承包含有内圈和滚动体的复合故障振动信号进行分析,结果表明该方法是有效的。

With the goodfeatures having multiple wavelet functions and scaling functions in multiwavelet, it cancommendably match various fault characteristics of multiple faults. Hence it has the capacity of extractingmulti-fault features simultaneously, a novel algorithm based on adaptive redundant lifting scheme multiwavelet packet transform was proposed.The optimization algorithm was applied to prediction operator and update operator, which minimized the signal residual Therefore, lifting wavelet basis can adaptively match observed signal The post-processing of output signal leads to extract the multi-fault synchronously which can put multi-fault into different channels.Experimental results show that the presentedalgorithm can effectively extract the multiple faults features of rolling element bearing in inner-race and rolling element