基于模糊Petri网的设备故障诊断新方法研究

专家系统用于计算机数控（CNC）设备的故障诊断日益受到人们的重视。为了能够解决故障现象的模糊性，本文将模糊理论与Petri网结合，建立了模糊Petri网（FPN）理论体系。引出了FPN转移被激发规则，确定了FPN的动态运行过程。应用FPN能够清楚地表达与或树、产生式规则等知识表达形式。基于FPN理论，建立了CNC设备故障诊断专家系统（CNCM-FDES）。该系统由知识库系统、知识库管理系统、诊断推理机制、诊断过程解释机制、故障评价、故障决策与入机界面组成。文中分别介绍了该系统的各个组成部分。

New Method Research for Machine Fault Diagnosis Based on Fuzzy Petri Net

Great importance is increasingly attached to expert system for fault diagnosis of computer numerical control (CNC) machines. In order to solve fuzziness of faults,fuzzy theory is connected with Petri net in this paper,and a fuzzy Petri net (FPN) theory is established. Rules firing transitions in FPN are introduced,and dynamic running process of FPN is determined. It is possible to clearly represent knowledge expression forms,such as and-or trees,production rules,etc. by FPN. Based on the FPN theory,an expert system for fault diagnosis of CNC machines (CNCM-FDES) is found. The system consists of knowledge base system, management system for knowledge base, diagnosis inference mechanism,diagnosis process interpretation mechanism, fault evaluation, fault decision and manmachine imerface. Each component part is separately introduced.