| 基于多群体遗传算法的非线性最小二乘估计 |
| |
| 引用本文： | 刘德玲,马志强. 基于多群体遗传算法的非线性最小二乘估计[J]. 东北师大学报(自然科学版), 2011, 43(1): 40-47 |
| |
| 作者姓名： | 刘德玲 马志强 |
| |
| 作者单位： | 广东食品药品职业学院,广东,广州,510520;东北师范大学计算机科学与信息技术学院,吉林,长春,130117 |
| |
| 基金项目： | 吉林省自然科学基金资助项目(20101506); 2009年广东省高等学校计算机基础课程教改与实践项目(A08) |
| |
| 摘 要： | 由于非线性模型参数估计理论广泛使用的传统牛顿类算法对初值的敏感性,以及简单遗传算法易陷入局部最优的问题,提出了一种多群体遗传算法,它采用多个群体执行遗传算法搜索解,并且能根据各个群体在较少迭代次数中找到的最优解动态调整参数域,提高了遗传算法的性能及搜索到的解是全局最优解的可靠性.实验结果表明:新的算法是一种有效的非线性...
|
| 关 键 词： | 多群体遗传算法 非线性最小二乘 参数估计 |
Nonlinear least squares estimation based on multiple genetic algorithms |
| |
| LIU De-ling,MA Zhi-qiang. Nonlinear least squares estimation based on multiple genetic algorithms[J]. Journal of Northeast Normal University (Natural Science Edition), 2011, 43(1): 40-47 |
| |
| Authors: | LIU De-ling MA Zhi-qiang |
| |
| Affiliation: | LIU De-ling1,MA Zhi-qiang2(1.Guangdong Food and Drug Vocational College,Guangzhou 510520,China,2.College of Computer Science and Information Technology,Northeast Normal University,Changchun 130117,China) |
| |
| Abstract: | Conventional Newton-like algorithms,widely used for parameter estimation of nonlinear models,are sensitive to initial values while simple genetic algorithms are liable to fall into local optimization.This paper proposes a multiple genetic algorithm.It searches the solution with several genetic algorithms and can adjust the parameter domain dynamically according to the optimum solution found by each genetic algorithm with several iterations,for which it can avoid running into local optimization,increase the ... |
| |
| Keywords: | multiple genetic algorithm nonlinear least squares parameter estimation |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
