多胞不确定时滞系统的输出反馈鲁棒预测控制

针对一类具有多胞结构的不确定离散时滞系统,提出了一种基于LMI的输出反馈鲁棒预测控制算法。根据Lyapunov稳定性理论,给出了保证系统鲁棒稳定性的充要条件,并将系统的鲁棒预测控制优化问题转化为易于求解的LMI问题。通过离线计算鲁棒预测控制器中的输出反馈增益矩阵,显著减少了算法的在线计算量。对系统输入、输出约束和状态时滞的考虑使所提出的鲁棒预测控制算法更接近于工程实践。而输出反馈的采用突破了以往算法中要求系统状态必须可测的限制,使算法具有更低的保守性。仿真结果验证了算法的有效性。

Output Feedback Robust Predictive Control for Polytopic Uncertain Time-Delay Systems

For a class of uncertain discrete time-delay systems with polytopic structure, an output feedback robust modelpredictive control algorithm based on LMI was proposed. According to Lyapunov stability theory, necessary and sufficientcondition was given to guarantee the robust stability of the system, and the robust predictive control optimization problemwas transformed into LMI problem which is easy to solve. The amount of online computation was significantly reduced bycalculating the output feedback gain matrix of robust predictive controller offline. The proposed robust predictive controlalgorithm is close to engineering practice because of the consideration of input and output constraints and state delay. Inprevious methods, the system state must be measurable, but the proposed algorithm broke through this restriction, so the al?gorithm has a lower conservative. Simulation results showed the effectiveness of the algorithm.