离散时间系统LQ最优控制逆问题解的数值算法  

Numerical algorithms for the weighting matrix Q of discrete time LQ optimal control systems

作　　者：任智华[1] 王耀青[2] 

机构地区：[1]天津工业大学计算机技术与自动化学院,天津300160 [2]武汉科技大学,武汉430081

出　　处：《天津工业大学学报》2001年第3期35-38,共4页Journal of Tiangong University

摘　　要：利用 L Q最优控制逆问题的参数化解 ,将求解对称、非负定加权矩阵 Q的问题变为一类 F-范数优化问题 ,给出了一种求解离散时间系数 L Q最优控制加权矩阵 Q的数值解法 .算法的优点在于任意给定一组自变量 ,通过解这类优化问题就可以求得满足闭环特征值要求的加权矩阵 Q。The numerical algorithms are given to solve the weighting matrix Q of discrete time systems by transforming the problem of determing a symmetric,nonnegative weighting matrix to a F norm optimization problem.The methods have the advantage that the weighting matrix can be obtained to meet the requirements of the closed loop eigenvalues by optimizing the F norm with a set of initial variables arbitrarily chosen.Simulation shows the algorithms to have good performance in convergence.

关 键 词：LQ逆问题 最优控制 加权矩阵 优化 数值算法 离散时间系统 

分 类 号：TP273.1[自动化与计算机技术—检测技术与自动化装置]