Prototype Design Of Vechicle-fault Diagnosis Expert System Based On Integration Of Case-based And Rule-based Reasoning

Due to the increasing complexity and risk of modern control systems,and the growing demands for quality,safety,cost efficiency and reliability,the call for fault diagnosis in automotive and their systems is gaining more and more importance. This paper introduces the trends of vehicle-fault diagnosis and describes some experiments performed on the application of case retrieval nets,which its merits were found to exceed what it could have achieved with rules or models,to the retrieval problem while building vehicle-fault diagnosis expert system based on case-based reasoning. Amelioration on original case-retrieval methods is also presented for producing more performance gains and the implementation of case retrieval nets is fulfilled by a relational database management system. Architecture is presented for integrating case-based reasoning and rule-based reasoning with totally new viewpoints. Moreover,this paper puts forward the framework of the expert system and the approach to implement the design above. The experiments are done in the domain of automobile sounder faults' diagnosis with the RAD tools (Borland Delphi5.0 and InterBase6.0),and the preliminary results are very promising.