Study on integrated intelligent system for fault diagnosis

This paper designs an integrated intelligent system for fault diagnosis that integrates multi-methods of fault diagnosis including rule-based, case-based, model-based and artificial neural network-based methods. The domain knowledge is represented in multi methods; the inference strategy and knowledge acquisition mechanism is also integrated. The study indicates that the integration of multi-fault diagnosis methods is an effective approach to overcome the limitations of single method, improve diagnosis efficiency, enhance diagnosis accuracy and extend covering ratio of the diagnosis results.