Formation Control of Multiple Mobile Robots Based on Iterative Learning Distributed Model Predictive Control

This paper proposes an iterative learning distributed model predictive control (ILDMPC) to control the formation of multiple mobile robots under uncertainty. Specifically, we design a general performance index constructed from the system's state variables and coupling parameters to replace the traditional cost function, promoting the system's control efficiency and the ability to seek the optimal solution. Furthermore, when dealing with and calculating the robot information, such as state variables and coupling parameters, information from the previous iteration is employed to construct closed-loop constraints in the optimization problem. Then, the results of the next iteration are calculated by learning the previous optimization trajectory and improving the overall system performance. Under the closed-loop constraints of the optimization problem, we analyze our system's feasible solution and iterative performance, demonstrating its effectiveness through several simulation experiments.