卷积型小波变换实现及机械早期故障诊断应用

Realization of convolution wavelet transformation and its application in mechanical incipient fault diagnosis

为消除Mallat算法存在的频率折叠等固有缺陷对机械故障诊断的不利影响,提出采用卷积型小波变换进行机械故障诊断。推导卷积型小波变换快速分解算法,给出基于滤波器组的递归分解实现方法;针对滚动轴承早期故障诊断与定量识别难题及共振解调法与冲击脉冲法的不足,提出将卷积型小波变换与共振解调法、冲击脉冲法相结合的新方法对滚动轴承早期故障进行诊断与定量识别,给出具体实现过程。仿真实验与实例分析表明:卷积型小波变换能消除Mallat算法固有缺陷对机械故障诊断的不利影响,较内积型小波变换更适合机械故障诊断。该方法可有效对滚动轴承早期故障诊断与定量识别,具有一定应用价值。

In order to avoid the influence of instinct defects in Mallat algorithm on mechanical fault diagnosis, the convolution wavelet transformation was proposed to conduct mechanical fault diagnosis. Then, the fast decomposition algorithm of the convolution wavelet transformation was induced in detail and a method based on filter banks to realize recursively the fast decomposition algorithm was presented. Aiming at the problem of diagnosis and identification of rolling bearing incipient faults and the shortcomings of the resonance demodulation and the shock pulse method, a new method was presented, it combined the convolution wavelet transformation with the resonance demodulation and the shock pulse method. Simulation and test analysis of bearing incipient fault data showed that the convolution wavelet transformation is fitter for mechanical fault diagnosis, and the new method can effectively detect the incipient faults and exactly identify weak damages of rolling bearing.