Research on Control of Quadrotor UAV Based on Improved BP Neural Network PID and ADRC

To address the challenges faced by quadrotor UAVs operating under dynamic and complex disturbance environments, this paper introduces an innovative flight control approach that synergizes an enhanced Backpropagation Neural Network (BPNN) with PID control and Active Disturbance Rejection Control (ADRC). Firstly, considering the disturbances during flight, the quadrotor model is established, and the position and attitude are controlled by PID and ADRC controllers respectively. Secondly, addressing the challenge of intricate controller parameter configuration, an enhanced Back Propagation (BP) neural network is employed to optimally adjust the PID controller parameters and the parameters of the Extended State Observer (ESO). Finally, the effectiveness of the outer-loop BPNN-PID and inner-loop BPNN-ADRC control method (BPNN-PID+ADRC) is verified by simulation experiments, and compared with the traditional PID and ADRC (PID+ADRC) flight control method. The results demonstrate that the designed method has enhanced adaptability and capability in rejecting disturbances.