A Method of Multi-echelon Facility Location Based on Genetic Simulated Annealing Algorithm

Reverse logistics network is the foundation for the efficient operation of reverse logistics system, and facility location is the core of the design of reverse logistics network. So, this paper develops a multi-echelon facility location model to design the best return network linking return points, return centers and manufacturing facilities. According to the character of the model, it proposes a solution based on genetic simulated annealing algorithm, in which, each chromosome consists of binary values and decimal values, and the special evolution operation is based on Metropolis rule. In order to satisfy all constraints, two sub-distribution algorithms for dealing with customer-return points and return points-return centers are designed. Finally, satisfactory facility location result can be obtained by simulation experiment. The result shows that the model and algorithm are effective and will have a preferable application future.