Research On Discrete Electromagnetism-like Mechanism For Assembly Sequences Planning

Assembly Sequence Planning (ASP) problem is one of the major problem considered in assembly technology. A good ASP program may reduce the assembly conflict in the process of production in effect, improve the production efficiency, increase the adaptability to the changes of the market, and help the company to fit the market requirements for increasingly diverse better. In this study, the Discrete Electromagnetism-like Mechanism (DEM) algorithm is used to solve the ASP problem.Firstly, the ASP problem and the EM algorithm are generally presented.Secondly, a connector-based ASP is presented. The ASP problem belongs to the combinatorial optimization problem, whose degree of complexity will higher with the increase of the information levels. So we need more time to find out the feasible and optimal solutions, solving the ASP problem with higher degree of complexity. In order to reduce the degree of complexity, a connector-based conception is used in the modeling process of ASP problem, and the degree of complexity is effectively reduced.Then, the theory of EM algorithm is presented, and DEM algorithm is proposed. The EM algorithm is a new population-based global optimization method. The algorithm simulates the attraction-repulsion mechanism in electromagnetism to move sample points toward optimality and is thus referred to as electromagnetism-like mechanism algorithm. Traditional EM algorithm is not suitable for ASP problem, as the ASP problem belongs to the discrete problems. In this paper, we study to use the Discrete Electromagnetism-like Mechanism algorithm to solve ASP problem. By redefining the distance and movement of solutions, initiating the memoriable mechanism, improving the calculation of charge and force, the electromagnetism-like mechanism algorithm pertaining to consecutive optimization problems is discretized and used to solve the assembly sequence planning problem. The results of experiments shows, electromagnetism-like mechanism algorithm is superior on the accuracy of solutions and costs less time in solving when compare to Guided genetic algorithm (Guided-GA), memetic algorithms and artificial immune systems algorithm.Moreover, three examples are calculated by the DEM algorithm, and the validity is verified by comparison. Finally, the full paper is summarized.