| With the continuous development of automobile industry,the problem of vehicle reliability follows.Improving vehicle reliability requires not only investment in the design and manufacturing stage,but also vehicle reliability test.Based on this test,a reliability growth model is established to guide the improvement of vehicle design.Taking the reliability of a vehicle as the research object,this dissertation analyzes the fault level,fault system and fault mode of the vehicle driving reliability test data.Based on the reliability growth theory and traditional model,the vehicle immediate correction reliability growth model and the vehicle reliability growth model with delay correction are established,and verified with the actual reliability driving test data of the vehicle.The specific research work is as follows:Firstly,the concepts of reliability growth process,reliability growth management and reliability growth model are described.The commonly used Duane,AMSAA and Gompertz reliability growth models are compared,and their advantages and disadvantages are analyzed.Duane model is selected as the vehicle reliability growth management analysis model,and AMSAA model is used as the basic model of single vehicle and multiple vehicle reliability growth models.Secondly,combined with the reliability test data of the three stages of the vehicle,the management of the vehicle reliability growth is analyzed.According to the test data,the fault location and fault type of the vehicle are analyzed.It is found that after the improvement of the first two stages,the number of various failure modes on the sample vehicle in the third stage is significantly reduced,the reliability is significantly increased,and the number of different failure modes is relatively balanced.Count the number of failures occurred in different mileage intervals,and analyze the vehicle reliability growth management through Duane model.The analysis results show that the overall reliability of the vehicle shows an increasing trend,but the reliability growth from the first stage to the second stage is slow,and the reliability growth in the third stage is obvious.Then,based on different test environments of vehicles,the concept of environmental factor is proposed,the reliability growth model of single vehicle immediate correction and AMSAA-BISE model of multiple vehicles are established,and two groups of test data are reserved to verify the model accuracy of the fault prediction value and fault prediction interval of the reliability growth model.The verification results show that the deviation between the real fault point and the fault prediction point is small,and the real fault point accurately falls within the fault prediction interval,The prediction accuracy of the model is high.The reliability growth of fault interval is analyzed by using the two groups of models.The analysis results show that with the progress of the test,the failure rate is significantly reduced,the reliability is significantly improved,and the reliability growth test is effective.Finally,combined with the environmental factor coefficient and correction effectiveness coefficient,the reliability growth model of vehicle with delayed correction is established,and the point estimation and confidence interval estimation of vehicle failure rate in three stages are carried out through the model.The calculation results show that the MTBF value in the third stage is significantly higher than that in the previous stage,and the corrective measures are effective,but the confidence interval is too large.On this basis,Bayes method is introduced to establish a multistage reliability growth model with delay correction,and the reliability growth is evaluated through the model.The research results show that this method adds a priori fault information,which can dynamically evaluate the multi-stage reliability growth process of vehicles,and solves the problems that the fault sample size decreases with the stages and there is no effective information transmission method between stages.The reliability growth model established by Bayes method has smaller prediction range,more accurate MTBF estimation,and more real and reasonable growth evaluation results of vehicles. |