基于线性矩阵不等式的分散鲁棒跟踪控制器设计

应用线性矩阵不等式（ＬＭＩ）方法研究不确定性关联大系统的分散鲁棒输出跟踪控制问题。系统中不确定项具有数值界，可不满足匹配条件。基于不确定基项的表达形式，给出了存在分散控制鲁棒跟踪控制器的ＬＭＩ条件。在此基础上，通过建立求解受ＬＭＩｓ约束的凸优化问题，提出了具有较小反馈增益ＬＭＩ设计方法，使受控系统渐近跟踪给定的参考输入。ＬＭＩ方法求解简单，最后用示例说明了该方法的应用。

The Design of Decentralized Robust Tracking Controllers Based on Linear Matrix Inequality

This paper studies the decentralized robust tracking control problem for large scale uncertain systems by linear matrix inequality (LMI)approach.The uncertainty is value bounded and can be dissatisfied with so called matching conditions.A sufficient condition for existence of decentralized robust tracking controller is derived.This condition is expressed as the solvability problem of linear matrix inequalities(LMIs).Based on that,a convex optimization problem with LMI constraints is formulated to design a decentralized state feedback control with smaller gain parameters which guarantees the controlled system to approach the reference input asymptotically.The given example shows the application of the method.