| 基于混沌变量的模式搜索法及其应用 |
| |
| 引用本文： | 袁小芳,王耀南,吴亮红.基于混沌变量的模式搜索法及其应用[J].湖南大学学报(自然科学版),2007,34(9):30-33. |
| |
| 作者姓名： | 袁小芳 王耀南 吴亮红 |
| |
| 作者单位： | 湖南大学,电气与信息工程学院,湖南,长沙,410082 |
| |
| 基金项目： | 国家自然科学基金
,
教育部高等学校博士学科点专项科研基金 |
| |
| 摘 要： | 提出了基于混沌变量的模式搜索法.在探索移动中,由混沌遍历性来生成移动方向和步长;在模式移动中,针对粗搜索与细搜索分别采取最优一维搜索和直接模式搜索.将全局优化能力强的混沌优化融入到模式搜索法中,且采取二级模式移动,搜索效率高.算法应用于模糊神经网络权值优化,仿真验证了其优良性能.
|
| 关 键 词： | 神经网络 混沌 模式搜索法 最优化 全局优化 |
| 文章编号： | 1000-2472（2007）09-0030-04 |
| 修稿时间： | 2006-12-02 |
Pattern Search Algorithm Using Chaos and Its Application |
| |
| YUAN Xiao-fang,WANG Yao-nan,WU Liang-hong.Pattern Search Algorithm Using Chaos and Its Application[J].Journal of Hunan University(Naturnal Science),2007,34(9):30-33. |
| |
| Authors: | YUAN Xiao-fang WANG Yao-nan WU Liang-hong |
| |
| Institution: | College of Electrical and Information Engineering, Hunan Univ, Changsha,Hunan 410082, China |
| |
| Abstract: | A new pattern search algorithm using chaos was proposed,in which the ergodicity of chaos was applied to produce stochastic move direction and length in exploratory move.Moreover,two-staged pattern move was employed for different search phases.The global search property of chaotic optimization was utilized in the proposed algorithm and it could realize global optimum,and two-staged pattern move had fast convergence velocity.The training of fuzzy neural networks was implemented by using the proposed optimization algorithm,and simulation results have demonstrated that it has efficient search ability. |
| |
| Keywords: | neural networks chaos pattern search algorithms optimization global optimization |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《湖南大学学报(自然科学版)》浏览原始摘要信息 |
| 点击此处可从《湖南大学学报(自然科学版)》下载免费的PDF全文 |
