运筹与管理 ›› 2018, Vol. 27 ›› Issue (3): 17-24.DOI: 10.12005/orms.2018.0055

• 理论分析与方法探讨 • 上一篇    下一篇

多目标0-1规划问题的元胞狼群优化算法研究

马龙1, 卢才武1, 顾清华1, 陈晓妮2   

  1. 1.西安建筑科技大学 管理学院,陕西 西安 710055;
    2.中软国际科技服务有限公司,陕西 西安 710077
  • 收稿日期:2017-05-08 出版日期:2018-03-25
  • 作者简介:马龙(1982-),男,博士研究生,讲师,研究方向:智能系统优化理论、矿业系统工程、算法设计等;卢才武(1965-),男,博士,教授,研究方向:系统优化理论、管理决策分析;陈晓妮(1980-),女,硕士,系统工程师,研究方向:智能优化理论、算法设计;通信作者:顾清华(1981-),男,博士,副教授,研究方向:系统优化理论、系统仿真等。
  • 基金资助:
    国家自然科学基金资助项目(51774228,51404182);陕西省自然科学基金资助项目(2017JM5043);陕西省教育厅专项计划项目(17JK0425)

Research on Solving Multi-objective 0-1 Programming by Cellular Wolf Pack Algorithm

MA Long1, LU Cai-wu1, GU Qing-hua1, CHEN Xiao-ni2   

  1. 1.School of Management, Xi’an University of Architecture and Technology, Xi’an 710055,China;
    2.Chinasoft International Technology Services Co.,Ltd, Xi’an 710077, China
  • Received:2017-05-08 Online:2018-03-25

摘要: 针对多目标0-1规划问题,首先基于元胞自动机原理和人工狼群智能算法,提出一种元胞狼群优化算法,该算法将元胞机的演化规则与嚎叫信息素更新规则、人工狼群更新规则进行组合,采用元胞及其邻居来增强搜索过程的多样性和分布性,使人工头狼在元胞空间搜索的过程中,增强了人工狼群算法的全局搜索能力,并获得更多的全局非劣解;其次结合多目标0-1规划模型对元胞狼群算法进行了详细的数学描述,定义了人工狼群搜索空间、移动算子、元胞演化规则和非劣解集更新规则,并给出了元胞狼群算法的具体实现步骤;最后通过MATLAB软件对3个典型的多目标0-1规划问题算例进行解算,并将解算结果与其它人工智能算法的结果进行比较,结果表明:元胞狼群算法在多目标0-1规划问题求解方面可获得更多的非劣解集和更优的非劣解,并具有较快的收敛速度和较好的全局寻优能力。

关键词: 狼群算法, 元胞自动机, 智能优化, 多目标, 0-1规划

Abstract: A cellular wolf pack algorithm,based on cellular automation principleand artificial intelligent wolf algorithm,is proposed for solving the multi-objective 0-1 programming problem. Using cellular and its neighbors to enhance the diversity and distribution of the search process,evolution rules of celluar automation and howling pheromone update rule as well as wolf update rule are combined, in order to realize global search and obtain more global non-dominated solution in cellular space search process. Secondly, search space, mobile operator, cellular evolution rules and update rules of optimal solution sets are defined by using mathematical form to describe the cellular wolfpackalgorithm. What’s more,the specific steps of implementation by celluar wolf pack algorithm are presented. Finally, the three typical examples of multi-objective 0-1 programming problem are the solution through the MATLAB software, and compared with the solutions of the other artificial intelligence algorithm,the results show that cellular wolf algorithm can get more non-dominated solution sets and better non-dominated solution in solving multi-objective 0-1 programming problem, and at the same time,the algorithm has faster convergence speed and better global search ability.

Key words: wolf pack algorithm, cellar Automation, intelligent optimization, multi-objective, 0-1 programming

中图分类号: