基于神经网络的智能故障诊断系统的开发研究

Development of a NeuralNetworkBased Intelligent Fault Diagnosis System

针对传统专家系统在故障诊断领域应用中的局限性 ,提出了将神经网络理论与专家系统技术相结合的混合方法用于现代制造系统的故障诊断过程 ,并介绍了神经网络的智能故障诊断系统的体系结构 ,并对系统性能和构造中的一些关键技术 ,如神经网络构造、数据准备和网络训练等进行了探索和分析。

A hybrid method of the neural network and the expert system is proposed for the fault diagnosis of modern manufacturing systems to solve the limitation of traditional expert systems. The general architecture of the neuralnetworkbased intelligent fault diagnosis system is presented and the key technologies, such as neural network building, data preparing, neural network training and etc, are studied and analyzed. Finally, as an application, the capability of this system is analyzed.\;