Robust Optimal Formation Control of Heterogeneous Air-Ground Vehicles Under Communication Faults via Reinforcement Learning

This paper develops a robust optimal formation protocol for multiple vehicles subject to communication faults via the reinforcement learning method. The nonlinearities and external disturbances in the vehicle dynamics and the communication faults among heterogeneous team are considered and a data-based robust coordination protocol is developed using only the local interaction information between vehicles. Simulation results of heterogeneous air-ground vehicles achieving the position coordination and yaw angle synchronization are presented to validate the proposed approach.