Research Based On Genetic Algorithm For Assembly Sequence Planning

Assembly sequence planning is the core of digital product pre-assembly. It is the foundation of selecting assembly device and fixture, and the basis of realizing assembly procedure simulation. So assembly sequence planning is worthy of researching. The problem of assembly sequence planning is a kind of NP-hard optimization problems, this paper presents a method of solving assembly sequence planning based on genetic algorithm aiming at the difficulty of assembly sequence planning. Firstly the paper researches on the assembly model. The assembly model presented by this paper includes geometrical, topological, basic part and fixture part information, to large scale assembly, divides it to some subassemblies. To reduce the input from user, the paper also researches on the method of auto get the information on the assembly model. Secondly the paper uses genetic algorithm(GA) to generate and evaluate assembly sequence. The GA adopts the encode method of matrix. The paper selects and designs GA operators, and analyses the characters of the operators, then reasons out the model of schema survival. According to the general principles in assembly, the paper designs the fitness function for the GA, and set the weight of fitness function through tests. It also researches on the translational and rotational stability. The paper adopts evolution strategy based phase, and set the GA parameters through experiments. Thirdly, development of the prototype system of assembly sequence planning, based on the research of above. Finally, It gives two examples to testify the correctness of all the above.