Applying Particle-Based Meta-Heuristic to Multi-Objective Inventory Planning

• Main Authors: Chia-hung Kao, 高嘉宏
• Other Authors: Ching-shih Tsou
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/95001109508533897277

碩士 === 世新大學 === 資訊管理學研究所(含碩專班) === 94 === The main characteristic of today’s business is volatility. Under such an environment, many decisions should be carefully thought over from all aspects. Therefore, traditional inventory planning should be recast into a multi-objective optimization problem (MOP). In a MOP, how to find the non-dominated solutions is not an easy task. This paper tries to apply two meta-heuristic algorithms, Particle Swarm Optimization (PSO) and Electromagnetism-like Mechanism (EM), to approximate the Pareto-optimal front for a multi-objective inventory planning problem. Subsequently, an outranking method called technique for order preference by similarity to ideal solution (TOPSIS) is used to prioritize the non-dominated solutions for decision makers. In a word, a two-stage multi-objective decision analysis framework which consists of multi-objective particle swarm optimization (MOPSO), multi-objective electromagnetism-like mechanism (MOEM) and TOPSIS is presented to simultaneously find out the lot size and safety stock decisions in an (r,Q) inventory system. By adjusting the weights of various objectives, including minimization of the annual expected total relevant cost, annual expected frequency of stock-out occasions, and the annual expected number of stock outs, through an interactive interface, decision makers can obtain a solution they satisfied most.