一种新的求解动态连续优化的分层粒子群算法-控制与决策.pdfVIP

第28 卷 第10 期 控 制 与 决 策 2013 年 10 月 Vol. 28 No. 10 Control and Decision Oct. 2013 文章编号: 1001-0920 (2013) 10-1573-05 一种新的求解动态连续优化的分层粒子群算法 1 1 2 朱庆保 , 徐晓晴 , 朱世娟 (1. 南京师范大学a. 计算机科学与技术学院，b. 江苏省信息安全保密技术工程 中心，南京210097 ；2. 安庆师范学院计算机信息学院，安徽安庆246001) 摘 要: 为了高效求解动态连续优化问题, 提出一种分层粒子群优化算法. 该算法将动态函数定义域分成 个子空 间, 每个空间用一个粒子群作为第一层进行独立搜索, 个子空间的最优粒子再组成一个全局粒子群进行全局搜索, 以达到全局牵引的作用, 同时提出探测环境和响应环境的策略. 利用经典的动态函数对算法进行测试, 结果表明所提 出算法能够迅速适应环境变化和跟踪最优解的变化, 效果令人满意. 关键词: 粒子群算法；动态优化；连续优化；分层 中图分类号: TP301.6 文献标志码: A A new hierarchical PSO algorithm for solving dynamic and continuous optimization problems 1 1 2 ZHU Qing-bao , XU Xiao-qing , ZHU Shi-juan (1a. School of Computer Science and Technology，1b. Jiangsu Research Center of Information Security Privacy Technology ，Nanjing Normal University ，Nanjing 210097 ，China ；2. School of Computer Information，Anqing Normal University ，Anqing 246001 ，China ．Correspondent ：ZHU Qing-bao ，E-mail ：zhuqingbao@njnu.edu.cn) Abstract: In order to efficiently solve the dynamic and continuous optimization problems, a hierarchical PSO algorithm is proposed. The solution space of a dynamic optimization problem is divided into subspaces. One particle swarm is assigned to each subspace to search independently as the first layer. Then the best particles in the particle swarms as the second layer search the global domain so as to guide the other particles in each subspace. Moreover, an environment detection and response strategy is proposed. Numerical experiments on the classical dyn