This paper reviews model predictive control (MPC) and its wide applications to both single and multiple autonomous ground vehicles (AGVs). On one hand, MPC is a well-established optimal control method, which uses the predicted future information to optimize the control actions while explicitly considering constraints. On the other hand, AGVs are able to make forecasts and adapt their decisions in uncertain environments. Therefore, because of the nature of MPC and the requirements of AGVs, it is intuitive to apply MPC algorithms to AGVs. AGVs are interesting not only for considering them alone, which requires centralized control approaches, but also as groups of AGVs that interact and communicate with each other and have their own controller onboard. This calls for distributed control solutions. First, a short introduction into the basic theoretical background of centralized and distributed MPC is given. Then, it comprehensively reviews MPC applications for both single and multiple AGVs. Finally, the paper highlights existing issues and future research directions, which will promote the development of MPC schemes with high performance in AGVs.

This article develops a new platoon control strategy for heterogeneous connected vehicles (CVs) subject to time delays and external disturbances. Specifically, based on the third-order vehicle model, a novel platoon controller is developed by embedding the variable time headway (VTH) spacing policy and the nonlinear motion coupling interactions between CVs. Simultaneously, an integral sliding mode (ISM) controller is developed to resist the disturbances. Then, the condition of asymptotic stability for the CV platoon and the upper bound of communication delay are deduced by using the Lyapunov theorem. Also, the string stability is proved by using the infinity-norm method. Finally, extensive simulations and co-simulations are provided to show the validity of the developed controller. Moreover, experiments with intelligent micro vehicles are conducted further to validate the practical feasibility of the developed controller.

Robust model predictive control (MPC) is a well-known control technique for model-based control with constraints and uncertainties. In classic robust tube-based MPC approaches, an open-loop control sequence is computed via periodically solving an online nominal MPC problem, which requires prior model information and frequent access to onboard computational resources. In this article, we propose an efficient robust MPC solution based on receding horizon reinforcement learning, called r-LPC, for unknown nonlinear systems with state constraints and disturbances. The proposed r-LPC utilizes a Koopman operator-based prediction model obtained offline from precollected input–output datasets. Unlike classic tube-based MPC, in each prediction time interval of r-LPC, we use an actor–critic structure to learn a near-optimal feedback control policy rather than a control sequence. The resulting closed-loop control policy can be learned offline and deployed online or learned online in an asynchronous way. In the latter case, online learning can be activated whenever necessary; for instance, the safety constraint is violated with the deployed policy. The closed-loop recursive feasibility, robustness, and asymptotic stability are proven under function approximation errors of the actor–critic networks. Simulation and experimental results on two nonlinear systems with unknown dynamics and disturbances have demonstrated that our approach has better or comparable performance when compared with tube-based MPC and linear quadratic regulator, and outperforms a recently developed actor–critic learning approach.