首页 | 本学科首页   官方微博 | 高级检索  
     

求解约束优化问题的多成员人工蜂群算法
引用本文:王翔,郑建国. 求解约束优化问题的多成员人工蜂群算法[J]. 西安交通大学学报, 2012, 0(2): 38-44
作者姓名:王翔  郑建国
作者单位:东华大学旭日工商管理学院;郑州航空工业管理学院土木建筑工程学院
基金项目:国家自然科学基金资助项目(70971020)
摘    要:针对约束优化问题提出了一种多成员人工蜂群算法.新算法设计了一种多成员机制,增强了在可行域内的搜索能力.在进行选择操作时,允许拥有较优目标函数的不可行解战胜可行解,增强了种群的分散性;在处理等式约束时,引入一种约束放松程度从大到小变化的机制,充分利用了等式约束周围不可行解的信息.针对13个标准测试函数的仿真实验表明:当处理含有等式约束且可行域较小的问题g13和最优解位于可行域内部且可行域较大的问题g02时,与改进人工蜂群算法相比,新算法最优解的均值误差分别减小了76%和80%.

关 键 词:约束优化问题  优化  人工蜂群算法  分散性

A Multi-Member Artificial Bee Colony Algorithm for Constrained Optimization Problems
WANG Xiang,ZHENG Jianguo. A Multi-Member Artificial Bee Colony Algorithm for Constrained Optimization Problems[J]. Journal of Xi'an Jiaotong University, 2012, 0(2): 38-44
Authors:WANG Xiang  ZHENG Jianguo
Affiliation:1(1.Glorious Sun School of Business and Management,Donghua University,Shanghai 200051,China; 2.School of Civil Engineering,Zhengzhou Institute of Aeronautical Industry Management,Zhengzhou 450005,China)
Abstract:A new multi-member artificial bee colony algorithm is proposed for constrained optimization problems.A multi-member mechanism is introduced in the algorithm to enhance the search capabilities in the feasible region,and infeasible solutions with better values of objective function are permitted to conquer the feasible solutions with worse objective function in selecting operators so that the population can be diversified well.Moreover,a mechanism of constraint relaxation is introduced to increase the probability of finding feasible solutions in dealing with linear equality constraints.The new algorithm is tested on thirteen well-known test problems.Comparison results with the modified artificial bee colony algorithm show that the mean errors of the new algorithm are reduced by 76% and 80% in handling the problem g13 with equality constraints and a small feasible region,and the problem g02 with the optimum inside a big feasible region,respectively.
Keywords:constrained optimization problems  optimization  artificial bee colony algorithm  diversity
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号