FMS 故障诊断专家系统的研究与开发

Study and Development of Fault Diagnosis Expert System of FMS

目的为保证柔性制造系统（ＦＭＳ）的可靠运行，对其发生的故障进行快速诊断，研究开发ＦＭＳ故障诊断专家系统．方法根据ＦＭＳ故障的特点，综合运用模糊数学、专家系统、神经网络、数据库和机器学习等理论，解决专家系统中许多关键技术，其中包括：模糊知识表达、基于感官直观现象的模糊快速诊断推理、神经网络的臆向推理、知识库管理和知识自动获取等技术．结果经现场实际应用，该专家系统达到快速、准确诊断ＦＭＳ故障的要求．结论先进理论的引入，使专家系统具有较高的实用价值．

Aim In order to ensure the reliable running of the flexible manufacture system (FMS), the fault diagnosis expert system for FMS was studied and explored Methods According to the characteristics of the fault of FMS, the theories of the fuzzy mathematics, expert system, neural network, database, machine learning and so on were applied synthetically to solve the many key techniques such as the representati on of fuzzy knowledge, the rapid fuzzy diagnositic reasoning based on the information of sensation and observation, the conjectural reasoning based on neural network, the management of the knowledge based, the automatic acquistion of fault diagnosis knowledge Results It was proved through practical application that this system met excellently the demand of the quick and accurate fault diagnosis of FMS Conclusion This system has the high practical value owing to the application of many advanced theories 