提升小波包变换及在振动故障诊断中的应用

Lifting Wavelet Package Transform and Its Application in Engine Fault Diagnosis Based on Cylinder Head Vibration

对内燃机进行快速准确故障诊断,可以提高内燃机维修的时效性,减少停机时间。提出一种基于提升小波包变换的故障诊断方法。给出了内燃机故障诊断模型,对内燃机进行等角度采样,在线获取振动信号,实现信号获取与工作转速同步。给出了提升小波包变换公式,并给出了移频算法,选择合适的小波进行提升小波包变换,提取状态特征向量,分析了不同状态特征向量的差异。采用支持向量机作为分类器实现故障诊断。通过实例分析比较,验证了该故障诊断方法的有效性。

Fast and exact fault diagnosis for engine can enhance maintenance efficiency and shorten shutdown period. A fault diagnosis scheme method based on lifting wavelet package transform is presented. A model for engine fault diagnosis is given. The vibration signal is collected on-line through equal angle sampling from the engine cylinder head, so, the synchronization of signal collection and rotate speed is realized. The formula of lifting wavelet package transform and the frequency-shift arithmetic are introduced. The state vectors are gained with the help of lifting wavelet package transform. The distinction between different state vectors is analyzed. Fault diagnosis is realized by a classifier based on support vector machine. Through analyzing an example, the validity of the scheme is tested.