摘要
研究了用改进的遗传算法（简称ＧＡ）求解同时镇定一族线性定常广义系统的最优输出反馈控制律问题。在满足稳定性的条件下，将最优同时镇定转化为一个受约束的非线性最小化问题。引入了自适应机制和惩罚函数变换，对传统的ＧＡ进行改造，并用于受约束非线性问题的全局优化。计算结果和数值仿真说明ＧＡ是求解同时镇定问题的一种有效的数值方法。
The optimal simultaneous stabilization problem for a collection of linear time-invariant descriptor systems via output feedback control law calculated by an improved genetic algorithm is considered in this paper. Under the condition of stability, the optimal simultaneous stabilizing problem is transformed into a constrained nonlinear minimizing problem. Using GA to optimize the constrained nonlinear problem globally, traditional genetic algorithm is improved by introducing the mechanism of adaptive and the transformation of penalty function. The calculation results and numerical simulations show that genetic algorithm is an effective numerical method for simualtneous stabilization problem.
出处
《电机与控制学报》
EI
CSCD
2000年第4期218-222,共5页
Electric Machines and Control
关键词
同时镇定
遗传算法
广义系统
惩罚函数
反馈控制律
simultaneous stabilization
genetic algorithm
descriptor system
penalty function
