Research And Application Of Automobile Fault Diagnosis Based On Bayesian Network

Automobile engine is a complex mechanical electricity's equipment which is composed of engine,electricity and liquid. Engine is the motivity of Automobile. Many basic functions of automobile have direct and indirect relations with engine. The 40 percent of automobile diagnoses come from auto-engine. So it is very important to judge diagnosis location,find out diagnosis reason and put forward ways when the engine is not disintegrated.Uncertain problems are key problems in fault diagnosis fields, which are resulted by many important reasons, including complex diagnosis objects, limit test means and the inexact diagnosis knowledge, etc. Automobile engine is large equipment, which has complex,correlative relation exiting in its units, and the units are full of uncertain factors and information. Traditional fault diagnose can not satisfy people of the request of large complex equipments' reliability. We need to find out another fault diagnose technology which is fit for uncertain problems. Bayesian network based on Bayesian theory is the best method for solving uncertain problems now.The Failure-Symptom topology structure is used to construct Auto model of Bayesian network. For solving uncertain problems, the on-line learning method of Bayesian network structure which is based on Bayesian theory is proposed aiming at working principle and diagnostic character of auto-engine. The paper uses the on-line learning method to revise Failure-Symptom topology structure and conditional probabilities table of network model. Then comparing with the traditional method of Bayesian network, we validate that on-line learning method of Bayesian network for fault diagnosis has high veracity and reliability on the result of Automobile engine fault diagnosis. Based on the theory, the paper analyzes and designs Automobile fault diagram system and then develops system module using Microsoft Visual C++.