圆形区域内鲁棒极点配置优化  

Optimization of pole assignment in a circular region

作　　者：翟军昌[1,2] 高立群[2] 

机构地区：[1]渤海大学信息科学与技术学院,辽宁锦州121013 [2]东北大学信息科学与工程学院,辽宁沈阳110819

出　　处：《控制理论与应用》2015年第12期1647-1652,共6页Control Theory & Applications

基　　金：国家自然科学基金项目(61273155;61503041)资助~~

摘　　要：本文基于和声搜索算法将精确极点配置的思想推广到圆形区域极点配置,通过对圆域内极点全局优化的方法,实现圆域内鲁棒极点配置问题.首先,基于几何原理描述极点在圆形区域内的位置信息,确定了在圆形区域内动态选择极点的规则,可以保证算法在圆形区域内动态选取极点.然后,对圆形区域内的极点,基于精确鲁棒极点配置的思想,利用和声搜索算法优化摄动或不确定性的谱范数上界,从而得到一组允许闭环系统具有更大摄动或不确定性的极点及相应的状态反馈控制器.最后,通过实例进行仿真,仿真结果表明本文采用和声搜索算法全局优化动态选择极点方法得到的闭环系统具有更好的鲁棒性.The idea of exact robust pole assignment is extended to region poles assignment by means of global opti- mization for poles in the circular region based on Harmony Search algorithm. First, a geometric description method for the position information of poles within the circular region is proposed to formulate the rules for the dynamic selection of poles, so that the algorithm can select poles dynamically from the circular region. Then, for the circular region poles, the harmony search algorithm is employed to optimize the upper bound of spectral norm for the perturbation or the uncertainty based on the idea of exact pole assignment. By this way, a set of poles and the corresponding robust state feedback controllers can be obtained, which allow the closed-loop system to have a maximum allowable perturbation or uncertainty. Finally, simulation results of the application example demonstrate that the dynamic selection of circular region poles will provide a closed-loop system with higher robustness than the other methods.

关 键 词：和声搜索算法 极点配置 状态反馈 控制器 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]