Research On Engine Fault Diagnosis Of Gasoline Vehicle Based On Artificial Intelligence

As an indispensable traffic work in modern life,automobile improves people's quality of life.Meanwhile,automobile operation is closely related to traffic safety and even life.As the "heart" of automobile operation,engine is the core component of automobile operation and the main source of automobile failure,Based on a joint brand as an example,whether in a short occupation or all warranty of period,repair and claim data at the upper level in the industry,but each year warranty of cost associated with the engine needs more 10 million RMB,The warranty of cost account for 70% of totally cost,proportion accounts for 60% of the proportion of the whole vehicle maintenance,maintenance and professional automotive quality research data show that 20% of customers on the fuel economy,engine noise &vibration and about high maintenance costs.At the present stage,the competition of automobile enterprises is the competition of products and services.The level of automobile fault diagnosis directly affects the quality and service level of vehicles and is closely connected with the development of enterprises.Often the repair shop USES the maintenance personnel and some detection equipment to carry on the vehicle fault diagnosis,but because of the maintenance personnel's technical level,causes the fault diagnosis time and the quality to be uneven,the breakdown diagnosis consumes the manpower and the material resources to be far larger than the maintenance link.With the progress of science and technology,the rise of artificial intelligence,cloud computing,big data,mobile network and other technologies has brought new development ideas for automobile engine fault diagnosis.Intelligent fault diagnosis will improve the accuracy,reliability and timeliness of diagnosis.Therefore,the establishment of an intelligent fault diagnosis model for automobile engines is of great significance to vehicle safety and the development of automobile enterprises.This paper focuses on the fault diagnosis of gasoline engine based on artificial intelligence.Aiming at the limitation of single intelligent diagnosis method,the automatic information processing technology--multi-information fusion technology is introduced to make full use of data information from different aspects and optimize the combination of information,so as to dig out more and more effective fault information.Summarize,summarize and analyze the corresponding relationship between fault symptoms and fault phenomena,and use artificial intelligence technology to accurately depict themulti-information fusion diagnosis model of data layer,feature layer and decision layer.This paper designs a multi-information fusion vehicle fault diagnosis model,which includes data layer,feature layer and decision layer.Among them,the method can deal with nonlinear problems,have self-learning ability and fault tolerance ability,and can quickly make fault classification neural network to construct fusion diagnosis model of data layer.Support vector machine(SVM)algorithm,which is suitable for small sample wood decision making and has strong generalization ability,is used to construct the fusion diagnosis model of feature layer.The fusion diagnosis model of decision layer is constructed by using the evidence theory which is more advantageous in expressing uncertain problems,and the fusion results of data layer and feature layer are used as evidence to further improve the accuracy of fault diagnosis.Finally,based on the characteristics of this experiment with engine each sensor design and develop analog device,to simulate the engine the problems in the process of actual operation,to get petrol engine data real time status,and the model of gasoline engine fault diagnosis based on artificial intelligence data layer,feature layer,decision-making fusion of diagnostic tests,the fault diagnosis model is verified by the experiments,accuracy and timeliness.