| 遗传算法应用于系统在线辨识研究 |
| |
| 引用本文： | 杨旭东,张彤,张家余. 遗传算法应用于系统在线辨识研究[J]. 哈尔滨工业大学学报, 2000, 32(1): 102-104 |
| |
| 作者姓名： | 杨旭东 张彤 张家余 |
| |
| 作者单位： | 哈尔滨工业大学,控制科学与工程系,黑龙江,哈尔滨,150001 |
| |
| 摘 要： | 研究了遗传算法在系统辨训识的应用，根据系统辨识自身的特点，设计了自适应选取遗传算法适值函数的方法，该方法的计算量要比排序选择操作的计算量小得多，在种群进化中采用自适应交叉、变异概率，以此对一差分方程描述的系统进行参数在线辨识。领导结果说明了该策略能够有效地避免算法非成熟收敛。
|
| 关 键 词： | 遗传算法 适值函数 在线辨识 系统辨识 仿真 |
| 文章编号： | 0367-6234(2000)01-0102-03 |
| 修稿时间： | 1997-06-19 |
Genetic algorithm for on-line system identification |
| |
| YANG Xu-dong,ZHANG Tong,ZHANG Jia-yu. Genetic algorithm for on-line system identification[J]. Journal of Harbin Institute of Technology, 2000, 32(1): 102-104 |
| |
| Authors: | YANG Xu-dong ZHANG Tong ZHANG Jia-yu |
| |
| Affiliation: | YANG Xu-dong,ZHANG Tong,ZHANG Jia-yu(Dept. Of Control Science and Engineering, Harbin Institute of Technology, Harbin 150001, China) |
| |
| Abstract: | Application of genetic algorithm to system identification is studied. According to the character of systemidentification, a new approach for adaptively choosing fitness function of genetic algorithrn is designed. Computationalamount of proposed approach is more less than that of ranking selection. In the process of optimization, adaptireprobabilities of crossover and mutation are adopted. Online parameter identification ?f a system described by differenceequations is done with a designed genetic algorithm. The simulation results show that this approach can preventpremature convergence efficiently. |
| |
| Keywords: | genetic algorithm fitness function premature convergence on-line identification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
