| 摘要: |
| 为了提高多参数寻优效果,基于人体生理双向协同网络调节机制,提出一种网络协同优化算法(NCEA)。对应相应生理系统设计其体系结构,具体包括监控管理级、协同调节级和群体搜索级:监控管理级根据适应度和群体分布密度等信息,发送协同指令给协同调节级;协同调节级根据协同调节指令,基于生理调节规律实时调整各个搜索群体的交叉和变异概率因子和搜索群体之间的个体交换概率,以及辅助群体的个体均匀化;群体搜索级包括主搜索群体和辅助群体,其中辅助群体为主群体提供优良个体,并避免搜索陷入局部最优。最后采用两个典型的多维非线性函数,检验NCEA的搜索精度和收敛速度,并应用于一种非线性智能优化控制器。试验结果表明,与标准的遗传算法和一种改进的遗传算法相比,NCEA具有较快的收敛速度和搜索精度。 |
| 关键词: 遗传算法 协同进化 交叉 变异 |
| DOI：10.3969/j.issn.1673-5005.2011.05.033 |
| 分类号::TP 274 |
| 文献标识码:A |
| 基金项目:国家自然科学基金重点项目(60534020);国家自然科学基金项目(60775052);中央高校基本科研业务费专项资金项目(09CX04026A);山东省优秀中青年科学家科研奖励基金项目(BS2010DX038);教育部博士点新教师基金项目(20070425518);中国石油大学(华东)博士科研启动基金项目(Y070502) |
|
| Collaborative evolution algorithm based on physiological bi regulation mechanism |
|
LIU Bao1, WANG Jun hong1, DING Yong sheng2,3, XIA Bo kai1
|
|
(1.College of Information and Control Engineering in China University of Petroleum, Qingdao 266555, China;2.College of Information Sciences and Technology, Donghua University, Shanghai 201620, China;3.Engineering Research Center of Digitized Textile ＆ Fashion Technology, Ministry of Education, Shanghai 201620, China)
|
| Abstract: |
| In order to improve the searching effect for the optimization of multi parameters, a networked collaborative evolution algorithm (NCEA) was presented based on the bi regulation mechanism of physiological system. Its structure was designed according to the relative physiological system, which included supper monitor level (SML), collaborative modulation level (CML), and searching population level (SPL). The SPL sends collaborative command to CML, according to the value of individuals fitness and distribution density fed back from searching population level. According to the collaborative command, the CML adjusts the crossover and variation probability, and the individuals exchange probability and uniformity of SPL based on the change of performance index and the corresponding law of physiological modulation. SPL is composed of main searching population and supplement searching population. The supplement population can supply excellent individuals for main population to avoid the searching falling into some local peak. In the experiments, two typical nonlinear functions were firstly selected to examine the searching precision and convergence rate of NECA, and then it was applied to the optimizing process of a novel nonlinear optimization intelligent controller. The experimental results show that the NCEA has better convergence rate and searching precision than normal GA and CGA (an improved GA). |
| Key words: genetic algorithm collaborative evolution crossover variation |
