基于蚁群算法的PID参数优化 | |
Alternative Title | Research and realization on parameters optimization of PID controller based on ant colony algorithm |
段海滨; 王道波; 黄向华; 朱家强; 北京8701信箱 | |
Department | 国家863计划微波遥感技术实验室 |
Source Publication | 武汉大学学报(工学版)
![]() |
2004 | |
Volume | 37Issue:5Pages:97-100 |
ISSN | 1671-8844 |
Language | 中文 |
Keyword | 蚁群算法 信息素 Pid 参数优化 |
Abstract | 针对传统的PID控制器参数多采用试验加试凑的方式由人工进行优化,提出了一种新型的基于蚁群算法的PID参数优化策略.蚁群算法是近几年优化领域中新出现的一种仿生进化算法,该算法采用分布式并行计算机制.在简要介绍蚁群算法基本思想的基础上,推导了蚁群算法PID参数优化方法,并给出了新算法的具体实现步骤,最后将该优化方案应用于某型高精度飞行仿真伺服系统.仿真应用研究表明,该PID参数优化策略具有很强的灵活性、适应性和鲁棒性,进而验证了该方案的可行性和有效性. |
Other Abstract | In light of traditional PID controller parameters optimization with manual cut-and-try method, a novel kind of PID parameters optimization strategy based on ACA(Ant Colony Algorithm) was proposed. ACA is a new category of bionic algorithm for optimization problems. Parallel computation mechanism is adopted in this algorithm. On the basis of brief introduction of ACA, a method for setting PID controller parameters using ACA was derived. A detailed realizing process of the new algorithm was also presented.In the end , this new PID parameters optimization scheme was applied to some high pre - cision flight simulation servo system . The simulation results show that the ACA based PID parameters optimization Iias excellent flexibility , adaptability and robustness . And the feasibility and effectiveness of this scheme 15 further verified . |
Funding Project | 中国科学院空间科学与应用研究中心 |
Document Type | 期刊论文 |
Identifier | http://ir.nssc.ac.cn/handle/122/1830 |
Collection | 微波遥感部 |
Corresponding Author | 北京8701信箱 |
Recommended Citation GB/T 7714 | 段海滨,王道波,黄向华,等. 基于蚁群算法的PID参数优化[J]. 武汉大学学报(工学版),2004,37(5):97-100. |
APA | 段海滨,王道波,黄向华,朱家强,&北京8701信箱.(2004).基于蚁群算法的PID参数优化.武汉大学学报(工学版),37(5),97-100. |
MLA | 段海滨,et al."基于蚁群算法的PID参数优化".武汉大学学报(工学版) 37.5(2004):97-100. |
Files in This Item: | Download All | |||||
File Name/Size | DocType | Version | Access | License | ||
200437597.pdf(179KB) | 期刊论文 | 出版稿 | 开放获取 | CC BY-NC-SA | View Download |
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment