人工神经网络和机械故障诊断

Artificial Neural Networks and Machinery Fault Diagnosis

智能化诊断是现代故障诊断技术发展的主要趋势,人工神经网络技术的出现为这种智能化提供了一个全新的途径。本文首先简单介绍了人工神经网络的基本性能及几个重要模型,着重探讨了人工神经网络技术在机械故障诊断领域中预测与控制、工况监测与故障分类诊断、模糊诊断和基于专家系统的故障诊断等几个主要方面的应用,指出人工神经网络技术与现有的信号处理、模式识别、模糊逻辑、专家系统等技术相结合,以解决故障信号分析与处理、故障模式识别以及故障论域专家知识的组织和推理等问题,必将加快智能化诊断发展的进程。可以预料:基于人工神经网络的故障诊断技术将具有广阔的发展与应用前景,并且随着VLsI 技术的发展,这一新技术必将广泛地应用于各种诊断实例。最后讨论了进一步值得研究的方向。

The development of modern fault diagnostic technologies is mainly towards the intellectualized diag-nostic technologies.Artificial neural networks'emergence provides a promising approach for this intellectu-alization.In this paper we first make a brief review on the fundamental properties of artificial neural net-works and several important models,and explore emphatically some potential application of artificial neuralnetworks in the area of machinery fault diagnosis:prediction and control,condition supervision and faultclassfication diagnosis,fuzzy diagnosis,and expert system based diagnosis.We point out that combining ar-tificial neural networks with signal processing,pattern recognition,fuzzy logic,expert system and othertechniques to the solution of the fault signal analysis and processing,the fault pattern recognition,organiza-tion and inference of the expert knowledge of the fault domain and so on,which accelerate surely the courseof intellectualized diagnosis.It is expected that the fault diagnostic technologies based on artificial neuralnetwork will have the broad development and application of perspectives.With the development of VLSItechniques,these new approachs will be applied extensively to various practical diagnostic cases.Finallysome directions deserving of further exploration are discussed.