Research On Generation And Optimization Of Assembly Sequence Based On Process Model

Feasible and optimal product assembly sequence has great influence on the improvement of assembly efficiency, the reduction of assembly cost and the improvement of assembly quality. Nowadays, with the more complexity and shorter development cycle of product, rapidly getting optimal product assembly sequence will become necessary to help enterprises to win in competitive market with shorter time, higher quality, lower cost and better service. Therefore, researching on assembly sequence rapid planning in digital environment has great theoretical significance and engineering application value.The goal of assembly sequence planning consists in generating feasible sequences and selecting optimal assembly sequence from the range of the feasible assembly sequence. Nowadays, the assembly sequence planning methods can be mainly classified into the traditional method and the soft computing method. They both have advantages and disadvantages. The former methods could effectively guarantee the feasibility of the geometric sequence by precise geometric reasoning. However, with the increasing number of the constituent part, the number of assembly sequence might grow quickly by geometry exponential and lead to composing explosion. The latter methods, especially the genetic algorithms, could generate the better result by using fuzzy knowledge. However, the quality and efficiency of its result accord to the quality and feasibility of the initial population.This thesis analyze the recent research of the assembly sequence planning, especially the advantages and disadvantages of the soft computing methods and the traditional methods. Try to integrate the advantages of two methods to establish more effective method to generate feasible and optimal assembly sequence to solve the problem of ASP. Based on this idea, this thesis firstly defines the product assembly process model, through the assembly state to express the assembly process, and establish the assembly state space. Secondly, analyze the interfere matrix and the relevancy matrix of relationship of assembly, and put forward three regulations to verify the validity and feasibility of assembly state, and design the matrix calculation formula, and validate the availability of the proposed method through the example. Research on the algorithms that applying graph search strategy to generate assembly sequence, analyze its shortcomings, establish an improved search algorithm and use example to illustrate the flow of applying the algorithm to generate assembly sequence. Analyze the weaknesses of genetic algorithm, research on the design means of objective function, define the concept of the stability of assembly state, and based on it to build the objective function. Finally, put forward an improved genetic algorithm based on improved search algorithm, and verify its correctness through the example.