第4期 刘宏达,等:均匀粒子群算法 341 HE Dakuo,WANG Fuli,ZHANG Chunmei.Establishment 5 结束语 of parameters of genetic algorithm based on uniform design 针对粒子群算法搜索效率随机,普通的粒子群 [J].Joural of Northeastem University:Natural Science, 算法难以满足某些实时优化的工程需要的现状,提 2003,24(05):409411. [7]方开泰,马长兴.正交与均匀试验设计[M].北京:科学 出了基于均匀设计法确定关键代次种群的S0算 出版社,2001:89-105. 法.利用均匀设计方法产生PS0算法的初始种群, [8]QIN Hong,ZHANG Shangli,FANG Kaitai.Constructing u- 使种群中的粒子在搜索空间分布更好地保持均匀分 niform designs with two-or three-level.Acta Mathematica 散性.提出和对比了4种均匀设计方案,且都没有增 Scientia,2006,26(3):451459. 加算法计算复杂性.通过测试和对比分析表明,新算 [9]ZHENG Yongling,MA Longhua,ZHANG liyan,QIAN Jix- 法能明显影响粒子的收敛性、快速性,能使算法具有 in.On the convergence analysis and parameter selection in 更稳定的搜索效率和搜索精度,减少粒子聚集和搜 particle swarm optimization[C]//Proceedings of the and In- 索早熟的发生 ternational Conference on Machine Learning and Cybernet- ics.Xi'an,China,2003:1802-1807. 参考文献: [10]孙先仿,范跃祖,宁文如.U均匀设计的均匀性研究 [1]KENNEDY J E R.Particle swarm optimization[C]//Proc [J].应用概率统计,2001,17(4):341-345, of IEEE Interational Conference on Neural Networks. SUN Xianfang,FAN Yaozu,NING Wenru.On the uni- Perth,WA,USA,1995:1942-1948. formity of U'uniform designs[J].Chinese Journal of ap- [2]刘宏达.粒子群算法的研究及其在船舶工程中的应用 plied probability and statistics,2001,17(4):341-345. [D].哈尔滨:哈尔滨工程大学,2007 [11]HICKERNELL F J.A generalized discrepancy and quadra- LIU Hongda.Research of particle swarm optimization algo- ture error bound J].Math Computation,1998,67:299 322. rithm andilts application in ship engineering[D].Harbin: Harbin Engineering University,2007. [12 ]FANG K T,QIN H.A note on construction of nearly uni- [3]张铃,张钹.佳点集遗传算法[J].计算机学报,2001,24 form De-signs with large number of runs[J].Statistics and (09):917-922. Probability Letters,2003,61(2):215-224. ZHANG Ling,ZHANG Bo.Good point set based genetic [13]Fang Kaitai.http://www.math.hkbu.edu.hk/Uniform algorithm[J].Chinese Journal of Computers,2001,24 Design/. (9):917-922. 作者简介: [4]XUE Guochen,XIN Li.Application of uniform design and 刘宏达,男,1976年生,副教授,主 genetic algorithm in optimization of reversed-phase chroma- 要研究方向为群智能理论及其应用,在 tographic separation[J].Chemometrics and Intelligent La- 群智能领域发表论文12篇. boratory Systems,2003,68(2):157-166. [5]薛明志,钟伟才,刘静,焦李成.正交Multi--Agent遗传算 法及其性能分析[J].控制与决策,2004,19(3):290- 294. XUE Mingzhi,ZHONG Weicai,LIU Jing,JIAO Licheng 马忠丽,女,1974年生,副教授,主 Orthogonal multi-Agent genetic algorithm and its perform- 要研究方向为机器视觉检测技术, ance analysis[J].Control and Decision,2004,19(3):290- 294 [6]何大阔,王福利,张春梅.基于均匀设计的遗传算法参数 设定[J].东北大学学报:自然科学版,2003,24(05): 409-411