Expert System for Fault Intelligence Diagnosis of Gasoline Engine

Abstract

This paper describes the composing of Fault Diagnosis Expert System of auto engine. A model for fault diagnosis expert system, based on artificial neural network and expert system, is proposed. Firstly, we build a diagnosis tree, which is based on a fault tree to build an expert system for Diagnosis, then, get training samples from the fault tree and combine the self-study function of ANN to analyze and diagnose faults from different aspects and layers by several different ways to improve the efficiency of system diagnose, overcome the disadvantage of traditional Fault Diagnosis Expert System. This system takes the Single chip microcomputer as a developing tool. It's well operated and visible. Compared with the results obtained by BP- ANN, our method has more fast convergence rate and high computation efficiency.It is an efficient and reliable novel fault diagnosis technology.