一种基于多目标遗传算法的非线性控制器  

A Nonlinear Controller Based on Multiobjective Genetic Algorithms

作　　者：李玉榕[1] 项国波[1] 

机构地区：[1]福州大学电气系,福建福州350002

出　　处：《计算机仿真》2004年第5期61-63,共3页Computer Simulation

基　　金：福州大学人才基金资助项目(XJB-0210)

摘　　要：该文利用相位滞后仅38.1°的Clegg积分器构成一个非线性比例积分器,并利用多目标遗传算法对该控制器的三个参数进行优化,其中被优化的两个目标分别为被控系统的给定性能指标和抗负载扰动能力。然后将该控制器应用于一个存在饱和特性和间隙特性的双闭环调速系统,研究并比较了该系统在阶跃给定输入下的性能指标以及抗负载扰动的能力。仿真试验表明将该非线性控制器应用于具有多个非线性特性的动态系统能取得优良的性能。In this paper, the Clegg integrator whose phase delay is only 38.1° is used to construct a nonlinear PI controller. And a multiobjective genetic algorithms is utilized to optimize the three parameters of the controller. The two selected objects are the performance of given input and the ability to resist the disturbance of load of the controlled system. Then the controller is used in a double closed loop speed-regulating system which contains characteristic of saturation and clearance. The performance of the system when its given input is a step signal and the ability to resist the disturbance of load are studied and compared. Simulation shows that the dynamic system which contains multi nonlinearity can possess excellent performance when the nonlinear controller is used.

关 键 词：积分器 多目标遗传算法 多目标优化控制 非线性控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]