状态时滞时变离散时间系统的最优预见控制器设计  

Design of an optimal preview controller for time-varying discrete-time systems with state time-delay

作　　者：廖福成[1] 徐玉洁[1,2] 

机构地区：[1]北京科技大学数理学院,北京100083 [2]北京科技大学自动化学院,北京100083

出　　处：《北京科技大学学报》2012年第2期211-216,共6页Journal of University of Science and Technology Beijing

基　　金：国家自然科学基金资助项目(No.61174209)

摘　　要：研究了一类具有状态时滞的时变离散时间系统的最优预见控制问题.所用的方法仍然是通过引入差分算子构造扩大误差系统.首先克服了差分算子不是线性算子的困难,成功构造了扩大误差系统.然后通过提升技术,把系统转化为形式上没有时滞的普通控制系统.最后通过引入可预见的目标值信号信息,得到最终的扩大误差系统.从这个扩大误差系统出发,利用时变系统最优控制的有关结果,设计处理原系统的带有预见作用的控制器.利用矩阵分解,把需要求解的高阶Riccati方程转化成一个低阶的Riccati方程.仿真实例表明了该设计方法的有效性.A type of optimal preview control problem for time-varying discrete-time systems with state time-delay is investigated.The methodology is to first construct an augmented error system by introducing a difference operator,and then transform the system into a traditional control system without time-delay in form by using the lifting technique to solve the problem that the operator is not a linear operator.Finally,preview information of the vector containing the desired tracking is introduced to obtain the ultimate augmented error system.Beginning with this augmented error system,a controller with a preview function which can process the original system is designed by making use of the results on optimal control for time-varying systems.Later,matrix partition is employed to transform the high-order Riccati equation that needs to be solved into a low-order Riccati equation.The effectiveness of the proposed design method is demonstrated by a numerical simulation.

关 键 词：状态时滞 时变系统 最优控制 预见控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]