Shi Ting, Shi Peng, Wu Zheng-Guang
IEEE Trans Cybern. 2022 Nov;52(11):11639-11648. doi: 10.1109/TCYB.2021.3078572. Epub 2022 Oct 17.
This article investigates the model predictive control (MPC) for discrete-time Markov jump systems (MJSs). First, the asynchronization between the modes of the controller and those of the plant is studied. An asynchronous MPC controller is designed to tackle this issue. Next, to reduce the computational cost and communication burden, a version of the dynamic event-triggered mechanism (ETM) is presented. Finally, the exogenous disturbances are considered and the notion of mean-square input-to-state stability (ISS) is taken into account in the controller design. The highlight of this article is the introduction of both dynamic ETM and asynchronous control into the MPC design. The control algorithm is developed and formulated as a convex optimization problem. Moreover, the recursive feasibility and the closed-loop mean-square ISS are both studied. Finally, some simulations are given to show the effectiveness of the derived MPC method.
本文研究离散时间马尔可夫跳跃系统（MJSs）的模型预测控制（MPC）。首先，研究了控制器模式与被控对象模式之间的异步性。设计了一种异步MPC控制器来解决这个问题。其次，为了降低计算成本和通信负担，提出了一种动态事件触发机制（ETM）。最后，考虑了外部干扰，并在控制器设计中引入了均方输入到状态稳定性（ISS）的概念。本文的亮点是将动态ETM和异步控制都引入到MPC设计中。所开发的控制算法被表述为一个凸优化问题。此外，还研究了递归可行性和闭环均方ISS。最后，给出了一些仿真结果以表明所推导的MPC方法的有效性。
