机械设备中黑箱部件的状态监测与故障诊断

Condition Monitoring and Fault Diagnosis of Black-Box Components of Mechanical Equipment

利用小波包分解、Yule- Walker AR谱密度分析算法和概率神经网络技术研究开发了一套状态监测和故障诊断系统 ,该系统是用于类似卷烟厂卷接包机八工位转塔的黑箱部件。利用仿真信号对系统的状态监测部分进行了测试 ,并应用到实践中去。在状态监测系统的基础上开发的基于概率神经网络的故障诊断系统 ,用仿真信号进行了测试 ,结果证明该系统是可行的。该系统的研制开发对类似黑箱部件的状态监测和故障诊断具有一定的实用价值 ,对其他类似机构的状态监测和故障诊断也具有参考意义。

A condition monitoring and fault diagnosis system is discussed in detail in this paper, which is applied to such a black-box structure as an eight-location-rotating-tower of a wrap-join-pack machine in a cigaratte factory. The condition monitoring subsystem is based on Yule-Walker AR spectrum analysis algorithm and wavelet packet analysis which is good at noise reduction. The fault diagnosis subsystem is based on probabilistic neural networks which are appropriate for diagnosis. The whole system is examined by simulated signals. The condition monitoring subsystem has been applied to field equipment. The subsystem is proved to be practical. The system is also beneficial to the condition monitoring and fault diagnosis of similar mechanical equipment.