Data-Driven Formation Control for Multiple Heterogeneous Vehicles in Air–Ground Coordination

This article addresses the data-driven robust optimal formation control of heterogeneous vehicles in air–ground coordination. The position and heading references for the quadrotor vehicles and the unmanned ground vehicles are generated through only the local information of themselves and their neighbors. Based on these generated references, a robust formation controller is constructed for the heterogeneous team to achieve the position formation with heading synchronization. Based on reinforcement learning theory, an optimal formation controller for the heterogeneous agents is learned without knowledge of the agent dynamics. Simulation results are presented to verify the effectiveness of the proposed formation control approach.