Rsearch On Intelligent Fault Diagnosis System Of Cnc Based On The Combination Of RBR And CBR

With the structure of CNC machine tools more complex, higher requirement is put forward on fault diagnosis and equipment maintenance, and merely relying on alarm message or fault reasoning can not meet the diagnostic needs of complex systems. Diagnostic systems with integration of a variety of reasoning technologies can give full play to the advantages of various technologies. It can improve the diagnostic efficiency of the system, and it is a trend of development of intelligent diagnosis.In the process of machine fault diagnosis, two diagnostic technology of RBR and CBR were used comprehensively, making the two technologies complement with each other in different stages of the diagnosis process, and an intelligent fault diagnosis system with integration of rule reasoning and case diagnosis technical advantages was established. This article introduced the construction process of the fault diagnosis system from the intelligent diagnosis system knowledge acquisition, knowledge representation and diagnosis reasoning.Firstly, we introduced various kinds of knowledge acquisition methods, and the method that extracts rules from the fault tree.Secondly, the paper analyzes common knowledge representation method, and respectively introduces the knowledge representation of alarm message, rules and case.Thirdly,two reasoning methods RBR and CBR were introduced. The former adopts depth-first knowledge search strategy and fault tree logic analysis to complete rule-based reasoning. The latter sets hierarchical index for case knowledge according to the structure of the machine, and in order to get similar cases , uses neighbor method for calculation of the similarity between current fault events and cases in case base. And on the basis of analyzing the relative merits of rule reasoning and examples diagnosis, intelligent fault diagnosis model based on RBR and CRB hybrid reasoning was proposed, which overcomes the shortcomings of RBR and CBR respectively and gives full play to the advantage of both.Finally, a prototype was developed with Qt, VS2005 and SQLITE,Which realizes the of fusion rule-based reasoning and case-based diagnosis. The system can not only make use of the field knowledge but can take advantage of diagnose experience in the past , which has a certain amount of diagnosis ability and improves the efficiency of the system diagnosis.