Research On Quality Control Methodology For Complex Product Manufacturing Processes Based On Data Variation Theory

In this paper it researches a quality control method for complex product manufacturing processes based on data variation theory,in view of the features of complex product components of a large number of parts,the interface relationship and quality characteristics variation in various ways,the main work of this paper includes:(1)It establishes a framework of quality control methods,a selection matrix of quality tools,a general standard system,and a quality variation measurement index system and the conversion relationship between some indexes for complex product manufacturing processes based on data variation theory;(2)Combined with the quality management requirements of aviation products,it propose a quality variation management process detailed as four stages of variation identification,variation diagnosis,variation optimization and variation control,and ten steps with working content,output results,as well as an application case on the manufacturing and assembly process of the 30th frame beam in an aircraft carrier platform;(3)It researches on methods and tools for identification and analysis of key characteristics,broke through the analysis technology of quality variation of complex product manufacturing processes based on the interpretative structural model,established the integrated framework and identification model,the data variation source identification process and method for multivariate nonlinear manufacturing processes based on the joint probability density,the likelihood function and likelihood comparison;(4)Based on no-failure probability,it present a reliability evaluation model,method and case study for multi-parameters and multi-processes technological system;This paper summarized the concept,analysis flow and calculation method of process capability index evaluation,and presented application cases of typical aviation product process capability index evaluation and analysis.Innovation points of this paper mainly include three aspects:(1)To create models and methods to identify the main variation source in manufacturing through multivariate nonlinear joint probability density function, likelihood function and likelihood comparison,this method solves the variation source diagnose problem that the traditional SPC technology can timely alarm but cannot diagnose variation source;(2)To propose a variation analysis technology of complex product manufacturing processes based on the interpretative structural model,by introducing the interpretative structural model to carry out hierarchical and classification of the prioritize,this method makes up the shortcoming of lack data supporting for the causality diagram,the fault tree and other quality tools.the model and algorithm of quantization in diagnosing the cause of quality problems can be used for the discovery and excavation of the deep level reason and logic relationships lead to quality problems;(3)To put forward a model and method of reliability assessment based on multi process technology system of multi parameter based on no-fault probability,this method can be used to carry out the process design,technical specification program,process improvement,technological system optimization.