Database Construction And Recall Warning Method Of Defective Automobile Products

Automobile defects will affect consumers,enterprises and society to varying degrees,slight defects will affect the use of car owners;serious defects will bring personal injury and a large number of property losses to consumers and enterprises,as well as serious social consequences.This also makes the relevant management departments and enterprises pay more and more attention to the management of automobile defective products,timely recall after finding defects,and reduce the loss and social impact caused by defects.Typical recall events of defective automobile products,such as "Takata airbag recall event","SAIC general shaft broken recall event",etc.From 2013 to 2017,the number of recalls of defective automobile products in China has been increasing year by year.At the same time,automobile defects and how producers respond in the event of a defect have increasingly attracted consumers’ attention.Therefore,in this paper,it is of great practical significance to use consumers’ first-hand data on complaints of defective automobile products to study a method that can accurately warn the manufacturer or the relevant management departments of defective automobile product recalls.This paper synthesizes domestic and foreign scholars’ research on early warning methods of defective automobile product recalls,and analyzes the theories of defective automobile product recalls and early warning and multiple linear regression to construct a defective automobiles product database based on consumers’ first-hand data on complaints about defective automobile products.A three-level multi-regression model of defective automobile product recall and early warning was established,and the reliability of the model was verified by statistical inspection.Finally,the model is applied through the complaints,consultations and recall data of defective automobile products,and the warning interval is optimized.The specific research work is as follows:1)Explain the background and significance of the research on the construction of defective automobile product database and recall early warning methods,summarize the domestic and foreign literature on defective automobile product recalls,analyze the research focuses and advantages and disadvantages of various scholars,summarize the research ideas of this paper,and define the research route.2)Utilizing the diversity of automobile defects,based on the analysis and extraction of the description of automobile defects,a four-part section was constructed,which included a catalog containing the common fault parts of defective automobiles,defective automobile failure modes,mileage and travel time of defective automobiles,and the seasons and regions where defective automobiles are located.3)According to the characteristics of data,based on the analysis of defective product theory,product recall theory and early warning theoretical knowledge,an early warning index system of defective automobile product recall including 8 indexes is constructed,and the division method of early warning interval of defective automobile product recall is studied.4)On the basis of the multiple linear regression function,a three-level early warning model of defective automobile product recall is established,namely,the key factor analysis model of the index,the single index impact recall analysis model and the full index impact recall analysis model.Regression testing,collinearity analysis,and residual distribution testing were performed to establish the full index impact recall analysis model to verify the reliability of the model.5)The recall model is used to calculate the actual case,and the accuracy of the model is analyzed.And further put forward suggestions for the recall management of defective automobile products.