Neural Network Modeling and Backstepping Control for Quadrotor

In this paper, a new approach based on Neural Network is proposed to identify the unmodeled dynamics for a quadrotor, and a nonlinear controller is adopted for the proposed model. First, the integrated model for a quadrotor is constructed with the dominant model and the compensatory model. The dominant model ref ects the quadrotor mechanical dynamics, while the compensatory model denotes the unmodeled dynamics. Then a nonlinear controller combining the PID approach and Backstepping approach is proposed for the integrated quadrotor model. The controllers of position, altitude and heading are designed to track the desired trajectory, while the attitude controller is used to track the desired angles produced by position controller. Finally, the effectiveness of the proposed model and the designed controller is demonstrated by a maze scenario based on the Airsim platform.