Formation cooperative trajectory tracking control for unmanned aerial vehicles via differential game and reinforcement learning

In this paper, a formation cooperative tracking control problem for unmanned aerial vehicles is investigated via non–zero-sum game theory and reinforcement learning. The formation cooperative tracking control problem is converted to a multi-player non–zero-sum game problem to attain the team optimality of multiple unmanned aerial vehicles. The desired trajectory can be achieved via a game-based feedback controller to minimize the global cost function involving all control policies. The Nash equilibrium control laws are learned based on only the collected system data, without knowledge of the vehicle dynamics. The asymptotic convergence of the constructed global closed-loop control system is proven via theoretical analysis. Simulation results validate the effectiveness of the proposed control method.