具有执行器饱和特性的不确定系统预测控制器设计

Design of model predictive controller for uncertain systems with actuator saturation

针对具有执行器饱和特征的不确定系统,提出了一种带有状态观测器的新型预测控制器设计方法.该方法在滚动优化的每一步,采用带有饱和特性的反馈控制结构得到一个最优控制律,使无穷时域性能指标最小.考虑在状态不完全已知的情况下,设计了带有状态观测器的预测控制器,并通过观测器参数调整使闭环系统渐近稳定.通过仿真实验验证了所设计控制器的有效性.

The design of novel model predictive controller with state observer is proposed for uncertain systems with actuator saturation. At each time step, the optimal control law that minimizes a infinite horizon objective function is obtained by applying a saturated feedback control structure. Then the observer is used when unknown state exists, and the model predictive controller with state observer is designed. The stability of close system is satisfied by adjusting the observer parameters. Finally, Simulation results show the effectiveness of model predictive controller with state observer.