基于状态观测器的约束鲁棒模型预测控制

Constrained Robust Model Predictive Control Based on State Estimator

模型不确定情况下的鲁棒问题是模型预测控制的一个根本问题。本文采用线性矩阵不等式(LMI),研究多模型不确定性描述情况下的鲁棒模型预测控制问题。在输入输出约束条件下,最小化最坏情况下的无穷时域目标函数,获得保证系统稳定的基于状态观测器的状态反馈增益并且给出观测器增益的设计方法。实例说明算法可行且保证闭环系统渐近稳定。

<Abstrcat>A fundamental question about model predictive control is robustness in conduction of model uncertainty. In this paper, the robust model predictive control system of polytopic uncertainty was studied by using linear matrix inequalities (LMI). The state feedback gain was acquired by minimizing an infinite horizon objective function based on state estimator,subject to constraints on the control input and plant output, and a designing approach to the estimator gain was given. and its feasibility of the controller design and the asymptotically stable of the closed-loop system is illustrated with a example.