Evolutionary Algorithms for Solving Unconstrained Multilevel Lot-Sizing Problem with Series Structure  

作　　者：韩毅 蔡建湖 IKOU Kaku 李延来 陈以增 唐加福 

机构地区：[1]College of Economics and Management,Zhejiang University of Technology [2]Research Center for Technology Innovation and Enterprise Internationalization,Zhejiang University of Technology [3]Department of Management Science and Engineering,Akita Prefectural University [4]School of Logistics,Southwest Jiaotong University [5]College of International Business and Management,Shanghai University [6]Institute of Systems Engineering,Northeastern University

出　　处：《Journal of Shanghai Jiaotong university(Science)》2012年第1期39-44,共6页上海交通大学学报（英文版）

基　　金：the National Natural Science Foundation of China(No.70971017);the Humanities and Social Sciences Project of Ministry of Education(No.10YJC630009);the Social Science Fund of Zhejiang Province(No.10CGGL21YBQ);the Natural Science Foundation of Zhejiang Province(No.Y1100854)

摘　　要：This paper presents a comparative study of evolutionary algorithms which are considered to be effective in solving the multilevel lot-sizing problem in material requirement planning(MRP)systems.Three evolutionary algorithms(simulated annealing(SA),particle swarm optimization(PSO)and genetic algorithm(GA))are provided.For evaluating the performances of algorithms,the distribution of total cost(objective function)and the average computational time are compared.As a result,both GA and PSO have better cost performances with lower average total costs and smaller standard deviations.When the scale of the multilevel lot-sizing problem becomes larger,PSO is of a shorter computational time.This paper presents a comparative study of evolutionary algorithms which are considered to be effective in solving the multilevel lot-sizing problem in material requirement planning （MRP） systems. Three evolutionary algorithms （simulated annealing （SA）, particle swarm optimization （PSO） and genetic algorithm （GA）） are provided. For evaluating the performances of algorithms, the distribution of total cost （objective function） and the average computational time are compared. As a result, both GA and PSO have better cost performances with lower average total costs and smaller standard deviations. When the scale of the multilevel lot-sizing problem becomes larger, PSO is of a shorter computational time.

关 键 词：simulated annealing(SA) GENETIC algorithm(GA) particle SWARM optimization(PSO) MULTILEVEL LOT-SIZING PROBLEM 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]