A Performance Prediction And Tolerance Grouping Combined Selective Assembly Method And Its Application

In the assembly process of precision products, due to its high assembly precision, complex matching relationships and complicated procedures, it generally has the problem of high fraction defective and low productivity. This article takes electro-hydraulic servo valve as study object, discusses its assembly performance prediction and tolerances grouping selective assembly method, and completes a performance prediction and tolerance grouping combined selective assembly system. This method has been successfully applied to improve the quality of assembly, reduce the rate of surplus parts, and improve the assembly production efficiency.Chapter1discusses the background and significance of the research, analyzes the problems of precise product assembly line, reviews about the tolerances grouping strategy, multi-object optimization selective assembly, assembly simulation and performance prediction at home and abroad. The final part of this chapter introduces the contents and chapter arrangement of this article.Chapter2discusses the prediction model construction strategy between the assembly influencing factors and performance indexes of electro-hydraulic servo valve. Based on the principle of multiple regression analysis, use record data from the assembly line to train the performance prediction model, and have the prediction model tested. To correct the error of the multiple regression linear prediction model, use neural network to modified the error of the prediction model, and improve the accuracy of the prediction model.Chapter3discusses a reality problem of the assembly performance prediction limitation, which is caused by the lack of the recorded data of unqualified products. Based on the mathematical model of electro-hydraulic servo valve construction and analysis, and also the analysis of assembly data distribution, a linear regression fitting model of the incomplete sample space is built, and the error of the fitting result is evaluated.Chapter4discusses the limitation of traditional tolerance grouping method in assembly performance optimization, and presents a performance prediction and tolerance grouping combined selective assembly strategy. Construct a precision product multi-object function based on Taguchi quality loss model, and use a hybrid particle swarm optimization as iteration algorithm. This method improves the assembly performance and tolerance grouping optimization at the same time, and improves the matching quality and product performance with the ensurance of pairing success rate.Chapter5discusses the construction of precision product selective assembly system. This system is designed and built based on the methods of performance prediction and tolerance grouping combined selective assembly and the analysis of the assembly process of electro-hydraulic servo valve. The system is applied in the assembly production of electro-hydraulic servo valve, and is verified of the validity and reliability in production.Chapter6summarizes the achievements of this article, and discusses the further research and some aspects that can be improved.