Attitude Synchronization for Multiple Quadrotors using Reinforcement Learning

In this paper, a reinforcement learning based control law is proposed to solve the attitude synchronization problem of the leader-following multi-quadrotor systems. The overall system is composed of a team of quadrotors, modeled with highly nonlinear and coupled dynamics. An optimal control solution is obtained by solving an augmented Hamilton-Jacobi-Bellman equation. A reinforcement learning approach is used to learn the optimal control law. Simulation results are provided to verify the effectiveness of the proposed controller.