Design Of Automobile Reliability Evaluation And After-sales Decision-making System Driven By Warranty Data

With the continuous growth of automobile ownership and the continuous construction of automobile data collection platform in our country,the automotive industry has accumulated a large amount of vehicle failure and usage data.Efficient data analysis methods and mining platforms are urgently needed to empower the development of the industry.Warranty data is data collected by the automobile after-sales service provider when the faulty vehicle is repaired,which is the most real reflection of the automobile quality.The in-deep mining of warranty data can provide important support for automobile reliability evaluation,enterprise after-sales decision-making and product design improvement.On this basis,this paper focuses on the reliability evaluation method,warranty cost prediction method and warranty strategy optimization method in warranty data analysis.Starting from the potential value of warranty data and considering the actual analysis needs of automobile enterprises,the warranty datadriven automobile reliability evaluation and after-sales decision-making system is built.The work and value of this paper are as follows:(1)Aiming at the problem that existing claim rate calculation model based on the n-MIS algorithm does not consider the maintenance situation of the automobile out of warranty will not be reflected in the warranty data,the claim rate calculation model based on the modification of the in-warranty rate is proposed.The model considers four situations of automobile products out of warranty during the two-dimensional warranty period,and corrects the total number of vehicles in the claim rate calculation model by using the in-warranty rate,which improves the calculation accuracy of the claim rate.On the basis,considering the difference in warranty costs between the two modes of part repair and replacement when the warranty of automobile products,the warranty cost analysis and prediction mode based on the number of claimed vehicles,repair-replacement rate and unit maintenance or replacement cost in the forecast period is proposed.Combined with the actual two-dimensional warranty data,the model is verified to accurately predict the warranty cost of automobile enterprises in the next quarter.(2)In order to improve consumer satisfaction and competitiveness of automobile enterprises,the warranty strategy optimization model considering consumer heterogeneity is proposed.Firstly,consumers are divided into two groups which are high-usage users and lowusage users,and the two-dimensional failure rate model is transformed into the one-dimensional failure rate model under different utilization rates.Secondly,four different preventive maintenance strategy and corresponding warranty cost calculation methods are proposed.Finally,aiming for keeping the total warranty cost of manufactures unchanged before and after the warranty strategy optimization,the basic two-dimensional warranty coverage of automobile products under different utilization rates is optimized by imposing preventive maintenance strategies on automobile products.The warranty strategy optimization model can reduce the failure rate of automobile products and improve user satisfaction.(3)Aiming at the functional requirements of automobile enterprises for warranty data analysis system,the warranty data-driven automobile reliability evaluation and after-sales decision-making system is designed.The data preprocessing method for warranty data and the database for supporting data analysis are designed in this system,and the reliability evaluation and repair spare parts prediction algorithm driven by warranty data is developed.Four functional modules of data management,statistical analysis,reliability evaluation and aftersales decision-making are built in turn,and the main interface of the system and the subinterfaces of each functional module are designed in detail.Combined with the actual twodimensional warranty data,the system is applied to execute fault statistical analysis,reliability evaluation and maintenance spare parts prediction.The proposed system can provide effective support for automobile reliability evaluation and enterprise after-sales decision-making.