Research Of Electronic Fuel Injection Engine Fault Self-Diagnosis Based On Neural Network Expert

The fault of engine is accurately diagnosed by the use of computer technology, neural networks and expert system. It is important to improve vehicle efficiency and quality of maintenance. With the widespread use of automobiles, research of automobile engine fault diagnosis system is of great significance.This paper takes fire-losing of the cylinder as diagnosis objects, introduces the basic theory and system structure of Artificial Neural Network and Expert System. The traditional Expert System only deal with the apparent knowledge, not only its inductive capacity is weak, but also the acquisition of knowledge has problems. The Artificial Neural Network offsets the shortage of Expert System. So it designs a new method which combines the Expert System with the Artificial Neural Network, this method can overcome each other's flaws. It makes application for fault diagnosis field better.Engine diagnosis Expert System is constructed based on Neural Network and combined with each situation in detail. The network is trained by the rule of the sample which was offered by Expert system. Then deal with diagnosis data through making use of the leaned network. At last, it obtains the diagnosis situation in detail.The detailed method is: it makes use of the Neural Network to learn rules, and it realizes the knowledge stored through applying BP Network, Finishing the knowledge construction and taking the inference, it realizes the man-machine interface through using VB 6.0, reaches the data output and input; it obtains a lot of diagnosis knowledge from books and the expert; using MATLAB to simulate the result, exerting the function of interpreter.The Neural Network Expert System as a system which applied to engine fault diagnosis, it has good practicality. So it can be expanded easily and expands the scope which diagnosis technique used in vehicle. In theory, it enriches the feasibility when vehicle takes the intellectual diagnosis, it makes engine fault diagnosis reach expert's level.