| Summary: | 碩士 === 國立臺灣科技大學 === 工業管理系 === 102 === Hierarchical facility location problem (HFLP) is an extended version of facility layout problem which aims to determine the location of facilities in a way that facilities in higher level can serve lower level efficiently. Particle swarm optimization (PSO) algorithms have been applied to a number of classical and real-world problems. However, since PSO is a stochastic search algorithm, it is likely to be failed to reach global value at the end of a run when the problem is too complicated. This study proposes a combination of PSO and TS for solving the hierarchical facility location problem (HFLP) called HPSO-TS. Moreover, this study also proposes the exact method branch and bound, genetic algorithm, and classical PSO to be compared to the proposed algorithm. To conform to the real world, this study considers the flow capacity in the networks or in other term the vehicle capacity, and in the results the optimal number of vehicle assignments also need to be considered. In this study, a mixed integer programming model for HFLP with flow capacity constraint is developed. The decisions to be made are locating a number of facilities on a set of potential sites, determining the network assignments, and number of flow assignments on the networks. The objective is to minimize the overall demand weighted distance travel, opening facilities cost, and flow assignments cost. A random key-based solution representation and decoding method is proposed for implementing HPSO-TS. The decoding method starts by transforming the particles into a priority list and then constructed as the assignment networks. Numerical results show that the proposed algorithm is effective and gives optimal or close to optimal solutions as compared with exact solutions in solving this kind of problems. Finally, the proposed algorithm is used to solve the study case of XYZ Company.
|
|---|
