Research On Fault Diagnosis For Electronic-controlled Engine-based On PNN Neural Network

With the continuous progress and development of science and technology,The electronic control system is more and more widely used in automobile.Experience method and fault codes in automobile fault diagnosis can not meet the actual needs.In order to find the fault part of the engine quickly and accurately,in order to improve the efficiency and accuracy of fault diagnosis,it is necessary to analyze and research the application of the right of the data stream.This article write the insufficient question of fault diagnosis in the engine only by experience method of fault diagnosis and fault codes,and analyse automobile data stream application situation analysis in the domestic auto repair industry,and summarize the application of data flow advantage and importance of fault diagnosis.In this paper,The Elantra 1.6L engine as an example,the PNN(probabilistic neural network)is applied to the fault diagnosis of the electronic control engine.First,The basic composition and the principle of self diagnosis of electronic controlled engine are introduced;The classification and analysis method of data stream is introduced in detail;.The difference between the traditional analysis method and the neural network analysis method is introduced in this paper.second,Characteristics and network structure of artificial neural network model and PNN are introduced in detail,through the characteristics of MATLAB,neural network fault diagnosis steps,PNN network design and MATLAB realization of detailed analysis and description,It provides a feasible basis for the fault diagnosis of the electronic controlled engine of the MATLAB software package;at last,The KT300 fault diagnosis instrument is used to detect the faults of the engine,They were built in a sample set.Through network learning on the sample data,inputing the data into the trained network program to verify the accuracy of fault diagnosis of neural network,and comparing PNN neural network diagnosis results with those by empirical analysis,It further illustrates the accuracy of using PNN network to diagnose the fault of engine data flow.The result of this paper provides a fast effective method for Automobile engine electronic control system fault diagnosis and detection,Improves the accuracy of fault diagnosis.