[1]曹国华,李婷婷.改进PSO算法及其在函数优化中的应用[J].南京师范大学学报(工程技术版),2007,07(02):010-13.
　Cao Guohua,Li Tingting.An Improved PSO Algorithm for Function Optimization[J].Journal of Nanjing Normal University(Engineering and Technology),2007,07(02):010-13.
点击复制

改进PSO算法及其在函数优化中的应用
分享到：

南京师范大学学报（工程技术版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
07卷
期数:
2007年02期
页码:
010-13
栏目:
出版日期:
2007-06-30

文章信息/Info

Title:
An Improved PSO Algorithm for Function Optimization
作者:
曹国华1;  李婷婷2
1. 南京师范大学电气与自动化工程学院, 江苏南京210042; 2. 镇江市丹徒职业教育中心, 江苏镇江212143
Author(s):
Cao Guohua 1; Li Tingting2
1.School of Electric and Automation Engineering,Nanjing Normal University,Nanjing 210042,China;2.Teaching Center in Dantu District,Zhenjiang 212143,China
关键词:
改进微粒群算法;  函数优化;  梯度法
Keywords:
im proved PSO;  function optim ization;  g rads a lgo rithm
分类号:
TP301.6
摘要:
针对微粒群算法PSO(Particle Swarm Optimization)应用于函数优化存在的问题,提出一种加入了梯度信息改进的微粒群算法.微粒群算法用于函数的优化,具有简单、效果好等优点.但是研究也表明该方法也存在着一些缺点,如计算时间较长、容易陷入局部最小等,这是由于算法本身的随机性决定的.梯度法是传统的优化方法,典型的特征是在确定优化解的方向时遵循梯度下降原则,因此在寻找优化值时方向比较确定,可以减少优化时间.为了克服PSO的缺点,在标准PSO优化策略中引入梯度原则,设计了一个具有梯度指导的PSO算法.现将改进的PSO算法应用于函数的优化,并与标准PSO算法的效果进行了比较.函数优化实验的结果表明,改进的PSO算法提高了标准PSO算法的收敛时间.
Abstract:
For the problem s appeared in the function optim ization o f the Particle Swarm Optim iza tion a lgo rithm ( PSO), an im proved PSO a lgor ithm attached by the grad inform ation is proposed in this pape r. The PSO a lgo rithm can be used in the problem s of function optim ization w ith characteristic o f simp lic ity, h igh e ffectiveness and so on. The prim ary study, however, show s that the op tim ization m ethod has som e sho rtcom ing s such as slow com puting speed, easiness to fa ll in local peak in la rge sca le prob lem, w hich is dete rm ined by the random ness of the algor ithm. The grad m ethod is a k ind o f trad itional optim ization m ethod and has the charac teristic that it is along the descend ing g rad direction o f optim iza tion va lues. So the grad me thod can reduce the tim e fo r the optim iza tion va lues because the d irec tion for optim a l values is determ ined by the g rad o f g rad algorithm. In o rder to overcom e the disadvantages of the standard PSO a lgor ithm, the princ ip le of grads m ethod w as inc luded in PSO a lgor ithm. Therefore, theG rads-PSO a-l go rithm ( regulated by grad m ethod) was proposed in this paper. The Grads-PSO algor ithm w as used in the optim ization o f function in this paper. The results obta ined by the Grads-PSO hav e been com pared by the ones o f the standard PSO a lgor ithm. The simu la tion resu lts show tha t the improved PSO a lgor ithm reduces the compu ting speed o f the standard PSO algor ithm.

参考文献/References:

[ 1] Kennedy J, Eberhart R. Pa rtic le sw arm optim ization[ C] / / IEEE Proceedings. IEEE In t Conf on Neura l Ne tw orks. Perth: IEEE, 1995: 1 492- 1 948.
[ 2] Ray T, L iew K M. A swarm w ith e ffectiv e info rma tion shar ing mechan ism for unconstra ined and constra ined single ob jec tive optim iza tion prob lem s[ C] / / IEEE Proceedings. IEEE Int Conf on Evo lutionary Com puta tion. Seou:l IEEE, 2001: 75- 80.
[ 3] Shi Yuhu,i Ebe rhartR. Param eter se lec tion in partic le sw arm optim ization[ C] / /Proc of the 7th Annual Con.f on Evo lu tionary Programm ing. W ashington DC, 1998: 591- 600.
[ 4] Eberha rt R, Shi Yuhu.i Tracking and optim izing dynam ic system s w ith pa rtic le sw arm s[ C ] / / IEEE Proceed ing s. IEEE Int Con f on Evo lu tionary Computation. H aw a i:i IEEE, 2001: 94- 100.
[ 5] 徐守江, 朱庆保. 基于微粒群算法的lp数据拟合及其应用[ J] . 南京师范大学学报: 工程技术版, 2006, 6( 3) : 62- 65.
Xu Shou jiang, Zhu Q ingbao. Partic le sw arm optim ization based on lp da ta fitting and its applica tions[ J]. Journal of N an jing No rm alUn iv ers ity: Eng inee ring and Techno logy Ed ition, 2006, 6( 3): 62- 65. ( in Ch inese)
[ 6] 吴献东, 金晓明, 徐志成, 等. 微粒群算法在模拟移动床色谱分离过程优化中的应用[ J]. 化工自动化及仪表, 2006, 33( 4): 5- 9.
W u X iandong, Jin X iaom ing, Xu Zh icheng, et a.l Application o f particle swarm optim iza tion in process of non- linear simu la ted m ov ing bed chroma tog raph ic frac tionato r [ J]. Contro l and Instrum ent in Chem ica l Industry, 2006, 33( 4) : 5- 9. ( in Ch inese)
[ 7] 吴亮红, 王耀南, 曾照福, 等. 基于复合微粒群算法的非线性系统模型参数估计[ J]. 系统仿真学报, 2006, 18( 7): 1 942 - 1 945.
W u L ianghong, W ang Yaonan, Zen Zhaofu, et a.l Pa rame ter estim ation o f nonlinear system s m ode l based on hyb rid partic les swarm optim ization a lgor ithm [ J]. Journa l of System S im ulation, 2006, 18( 7): 1 942- 1 945. ( in Chinese)
[ 8] 原萍, 陈红, 王光兴. Adhoc网络路由优化的微粒群方法[ J]. 小型微型计算机系统, 2006, 27( 7): 1193- 1196.
Yuan Ping, Chen Hong, W ang Guangx ing. Particle sw arm optim iza tion fo r routing design in Ad hoc ne tw orks [ J]. M in-iM icro System s, 2006, 27( 7): 1193- 1196. ( in Chinese)
[ 9] 陈国初, 俞金寿. 单纯形微粒群优化算法及其应用[ J]. 系统仿真学报, 2006, 18( 4): 862- 865.
Chen Guochu, Yu Jinshou. Sim plex pa rtic le swarm optim ization a lgor ithm and its applica tion [ J]. Journa l o f Sy stem S im ulation, 2006, 18( 4): 862- 865. ( in Ch inese)
[ 10] 沈洪远, 彭小奇, 王俊年, 等. 基于混沌序列的多峰函数微粒群寻优算法[ J]. 计算机工程与应用, 2006, 42( 7): 36- 38.
Shen H ongyuan, Peng X iaoq,i W ang Junnian, e t a.l A PSO algor ithm based on chaos sequence form ult-im odal function opt-i m ization [ J]. Compu ter Eng ineer ing and App lication, 2006, 42( 7): 36- 38. ( in Chinese)

备注/Memo

备注/Memo:
作者简介: 曹国华( 1963-) , 副教授, 主要从事检测技术与自动化装置等方面的教学与研究. E-m ail: cgh@ m ail.nsgk. net
更新日期/Last Update: 2013-04-29