Research On The Quality Defect Evaluation Of Vehicle Audit Based On Subjective And Objective Integration

The main purpose of vehicle audit is to evaluate vehicles in the factory which are ready to customers for delivery.Management can know about the quality level of vehicles for that.Although the audit content,defect classification and defect score evaluation of the auto production enterprises are different,feedback of quality defect durning vehicle audit is the types of score.In accordance with relevant standards,auditors inspect and feedbak quality defects based on customers' position.The evaluation result of vehicle audit is completely influenced by auditors and their comprehensive qualities.Therefore,in order to reduce auditors' subjective evaluation factors,the object of this thesis is to introduce subjective and objective fusion theory and analysis into vehicle audit.In order to research on quality defect evaluation of vehicle audit based on subjective and objective fusion technology,the thesis is to solve auditors' subjective evaluation of vehicle audit.Subjective and objective fusion of vehicle audit are respectively researched from the following aspects:(1)Monomer quality defect is the basic element of vehicle audit.According to basic principle of monomer quality defect evaluation,the thesis analyzes its characteristic and establishes a function model.Evaluation of defects is based on subjective evaluation,type and location of defects.Analyzing indicators and specific criteria of vehicle audit,quality defects are classified and arranged according to quality defect types of vehicle audit.Then a quality defect evaluation index system is set up based on layered architecture.The same type of defects will be classified to streamline operations on the basis of the layered architecture.According to actual audit situation and characteristics of quality defects,proposing that evaluation indexes should be divided into interval and membership shall be established.Then AHP method is used to calculate weight of index system.Finally,in accordance with layered architecture of index system and calculation of weight and membership,using fuzzy vector mapping technology method to establish quality defect evaluation system based on the layered architecture,determining evaluation grade range of dynamic threshold value,and finally realizing diversity and intelligent of evaluation results.(2)Combined related analysis with subjective index system of manipulation stability at home and abroad,a subjective evaluation index is built,and entropy value method is used to calculate index weights.Setting up a subjective evaluation score model to calculate subjective evaluation value of test vehicles.Choosing manipulation stability tests to establish objective indicators dual screening mechanism of vehicle handling stability performance.Prescreening objective index and correlation coefficient is obtained by using correlation analysis,and then objective evaluation parameters are downsized dimension on the basis of KPCA principle,finally first nuclear objective evaluation is extracted to reflect the manipulation stability.According to the score of subjective evaluation and objective evaluation parameters,improvement SVM method is estabilished to predict the interaction model which predicts the result of the subjective evaluation with less objective indexes.Improved artificial fish algorithm is used to optimize SVM parameter.Predicted value of objective parameters is obtained by the model which realizes subjective and objective evaluation fusion in the process of road audit test.(3)In accordance with analysis of traditional defects,improved FMEA technology is introduced to analyze the quality defects of vehicle audit.On the basis of vehicle audit characteristics,properties of S,O,D are respectively established to fit tor subjective evaluation principles of vehicle audit and suitable for objective calculation model based on defect data statistical,which realize vehicle audit quality defect characteristic corresponding to the attributed parameter.Using the rough C correlation matrix to weight the attributed parameters,the main objective fusion calculation model of RPN is reestablished which can solve accuracy and sensitivity of RPN values.Using the newly established RPN model,the three parameters are considered to carry out multi-attributed risk sequencing mechanism and analysis of quality defects,so as to avoid the problems that only focus on severity of defects.Based on SNR,multivariate quality loss function model is established,which minimums the quality defect loss as optimization goal.Through improved genetic algorithm,optimization is used to solve corrective action decision-making of quality defects.Subjective and objective fusion of quality defect analysis and corrective decision-making action are realized.(4)A quality defect evaluation system is developed based on vehicle audit.The system is composed of system design,system function,system evaluation and of system function realization.According to the needs of the audit task,the case implementation of the vehicle audit is carried out.The final evaluation results are consistent with the corresponding results of audit score,and the calculation is verified.The thesis' s researching results can be used in the process of actual audit work,solving a series of strong subjectivity problem of auditors which include the evaluation results,the suspected quality defects during road test audit,defects analysis and corrective decision-making action,providing theoretical basis for vehicle audit work improvement and significance to practical production.