Genetic simulated annealing algorithm-based assembly sequence planning

Assembly sequence planning plays an important role in the product development process. It is an important factor that determines quality and cost of the product assembly. Cost in assembly can be reduced by the implementation of generating automatic product assembly sequences, and selecting the optimum sequence in product assembly process. Assembly sequence planning (ASP) is combinatorial problem. Graph-based algorithms are adopted for traditional ASP method. In recent years, some genetic algorithms and simulated annealing algorithms have been used to solve ASP problems, and some achievements are arrived at. However, the two kinds of algorithms have limitations for ASP. GA heavily depends on the choosing original sequence, which can result in early convergence in iterative operation, lower searching efficiency in evolutionary process, and non-optimization of final result for global variable. For simulated annealing algorithms, the principle of generating new sequence is exchanging position of the randomly selected two parts. Obviously, for complex products, a number of non-feasible solutions may appear, and the efficiency is low. In view of these limitations, the approach of combining GA and SA is proposed to build genetic simulated annealing algorithm for the optimization of ASP. In this paper, the following contents are included. Firstly, the relevant researches on assembly sequence planning and the application of GA and SA are summarized. Next, the idea of combining the two algorithms into genetic simulated annealing algorithm is put forward, which aims at improving the efficiency of problem solving. Thirdly, the genetic simulated annealing algorithm for assembly sequence planning is implemented, the method, procedure as well as key techniques of the genetic simulated annealing algorithm are addressed in detail and the principles for selecting parameters are studied to achieve better performance of the algorithm. Fourthly, a case study is presented to validate th- proposed method. In the case, GA, SA and genetic simulated annealing algorithm are applied to ASP respectively, and the results verify the advantages of the genetic simulated annealing algorithm in solving the ASP problem. At last, the work of this paper is summarized and the future researches are given.