Adaptive finite-time tracking control of quadrotor UAV subject to input saturation

In this paper, a finite-time trajectory tracking control scheme is established for the quadrotor unmanned aerial vehicle (UAV) under the co-existence of modeling uncertainty, external disturbance and input constraint. Firstly, a deep deterministic policy gradient (DDPG) based auxiliary part is established estimate and compensate the model uncertainty and external disturbance. Then, design the sliding surface with tracking error for the backstepping control procedure and use Lyapunov stability theory to solve the optimal control law. Besides, the control law has a hyperbolic tangent function form which can helps to handle the input constraint problem. Finally, validate the effectiveness of the proposed control scheme through simulations.