Abstract:

Because of the limitations of a singleexpert system or neural network system in fault diagnosis, a new integrative fault diagnosis expert system which combines neural network technique with expert system is proposed. This system is applied in the mechanical breakdown diagnosis of CNC machine. The structure, characteristics and diagnostic methods of the neural network expert system are introduced. The system is verified based on the knowledge of machine failure and sampling data of faults. The result shows that this system has friendly interface and is easy to handle. It raises the level and efficiency of mechanical fault diagnosis of CNC machines.

Key words: vibration and wave, CNC machine, neural network, expert system, fault diagnosis

摘要：

依据单一的专家系统或神经网络在处理故障诊断中各自存在着局限性，提出将神经网络技术与专家系统融合的集成式故障诊断专家系统,并用于数控机床的机械故障诊断中。介绍神经网络专家系统结构、特点及诊断方法,利用获得的机床机械故障知识、故障样本数据对系统进行验证。结果表明该系统人机界面友好,操作简单,有效地提高数控机床机械故障诊断的水平和效率。

关键词: 振动与波, 数控机床, 神经网络, 专家系统, 故障诊断