Research Of Automobile Engine Performance Test And Fault Diagnosis Expert System

Automobile is a complicated mechanical system made of thousands of accessories. Engine is one of the most important components of automobile, and it is the power resources of automobile. Some basic technology performances of automobile are connected with correlative performances of engine directly or indirectly. Compared with developed countries, the difference between vehicle and road is great, and the culture degree of drivers and inspector is low. So engine performances test is greatly important to know the whole performance of automobile. In this paper, the automobile engine performance test and diagnosis expert system is studied. By means of this system, users can analyze and diagnose the faults fast and easily. As a result it can decrease the blindness of repair, improve the reliability and security o f automobile, shorten the repair time and cut the repair cost. So it has an important economic value and wide market. The expert system(ES) is based on symbol inference and it is an intelligent model. And the essence of it is a computer programme that the human experts use their knowledge to solve the problems in their fields. It can inference under the conditions that the system is unstable and the field knowledge is not self-contained. And can explain the process of problem solving. While it has some defaults such as the difficulty to getting the knowledge, the narrow range of knowledge and the weak inference. Artificial neural network (ANN) is a complicated network which is consisted of a lot of nerve cell. It is a dynamic system with the structure of vectorgraph. It can deal with the information from input. The value of the network can get from the training. It has the characters of study, induce and memory. But it has some shortcomings such as the results are not easy to be understood and explained. So the a model of fault diagnosis expert system based on neural network is proposed, which has the advantages of the neural network and traditional expert system. It make the quickly and accurate fault diagnosis to come true. It is a new method for engine fault diagnosis. The main part of this paper is automobile engine fault diagnosis system based on neural network and the hardware and software design for data collection system. The data collection system which is programmed by assemble language is an important part of fault diagnosis system. It consists of many sensors like some stress sensors and electric currentlvoltage sensors. The single chip processor 8901 deals with the signal getting from different sensors by AID, magnifying, and analyzes the principle of sensors choice and diagnosis parameters. Some measures against interfere are advanced. The pre-processing of data in engine抯 test is introduced. According to the knowledge that consists of the structure and principle of automobile engine and the experience from experts, the fault trees are established by using of computer technology. The characters of the ES take root ANN and integrated ANN. The train stylebooks comes from fault trees and provide them to the neural network for study. The value consists of one part of the repository. The ES can undertake self-learning to expand knowledge library and raise the system 憇 adaptability and error-allowablity, and make the fault diagnosis more accurate. The BP(Back-Propagation) network抯 programming design and arithmetic steps with the sigmoid function is illuminated exactly in this paper, and discuss the fuzzy fault diagnosis based on BP network. The system is designed with the oriented programming language 梀B 6.0 and ACCESS da...