Abstract:

Consider a constrained LPV system with polytopic description. An offline design of state observer is proposed when the system state cannot be observed. And the new self-tuning system constructed by state observer error is proved to be robust stable, which guarantees the output of the observer can converge to the true sate value. After that, the parameter dependent Lyapunov function (PDLF) is introduced to obtain poly-quadratically stable control laws by solving the min-max optimization problem with infinite horizon performance cost. And the output feedback control law can guarantee the feasibility and stability of the computational paradigms. Compared with the traditional strategy, the offline design method of observer can greatly reduce the online computation and the utilizing of PDLF can obtain the less conservative condition for the LPV systems. And at last, the simulation results show the effectiveness of the proposed method.

摘要：

针对多包描述的约束LPV系统，当系统状态不可观测时，给出一种离线状态观测器的设计方法，通过证明由观测器误差构成的自治系统是鲁棒稳定的，保证了观测器输出将最终收敛于系统状态真实值。基于状态观测器，对于系统的观测状态引入参数Lyapunov函数，通过求解无穷时域min-max优化问题给出鲁棒输出反馈预测控制律，并证明了算法的可行性和系统的闭环稳定性。与传统观测器设计方法相比，离线状态观测器设计方法有效的降低了在线计算量，同时采用参数Lyapunov函数的设计算法，在满足系统约束的情况下及稳定性的条件下减少了设计的保守性，并对不确定参数有更好的鲁棒性。仿真结果验证了算法的有效性。