摘要
为使PID参数整定工作更加快捷有效,在传统遗传算法的基础上,提出基于独特型免疫动力学网络模型的新型免疫算法,以激励水平作为抗体选择操作的量度,并构建精英抗体库,以便给精英抗体更多交叉变异的机会。将该算法用于PID控制器参数整定,仿真结果表明,新算法能更快地找到控制器PID参数最优组合,且经该算法优化后的系统控制性能优于传统遗传算法优化的PID控制器,故该算法用于PID参数优化是可行且有效的。
In order to make the PID parameters setting work quicker and effecter, a new immune evolutionary algorithm based on immune dynamical network model is proposed.The algorithm adopts selection operation according to the stimulation level of each antibody, and a memory base for good antibodies is devised so that more chances for them to transposition. The mutation results of the setting of PID parameters show that the new algorithm is utilized to quickly locate the optimal parameters of PID controller. And the performance of PID controller based on this new algorithm is superior to that of PID controller based on genetic algorithm. It is practical and active to apply the algorithm is proposed to PID parameter optimization.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第10期2461-2463,共3页
Computer Engineering and Design
关键词
PID控制器
人工免疫
遗传算法
独特型网络
参数优化
PID controller
artificial immune
genetic algorithm
idiographic network
parameter optimization
作者简介
司徒莹（1972-），女，广东开平人，硕士，讲师，研究方向为智能控制。E-mail：situy@163．com
