Adaptive neural network control for quadrotor unmanned aerial vehicles

Abstract

This paper studies the problem of adaptive attitude stabilization and position control for a quadrotor unmanned aerial vehicle (UAV) system with unknown variable payloads. Based on Back Propagation neural networks, an adaptive PID controller is proposed for quadrotor UAV with unknown variable payloads. The performance of the proposed adaptive PID controller is experimentally analyzed with comparison to classical PID controller. It is shown that the neural network adaptive PID controller is capable of dealing with the unknown variable payload issue by adjusting the PID parameters online. Simulation results confirm that the proposed neural network adaptive PID controller outperforms the parallel PID controller.