Trajectory Tracking of a Quadrotor based on Gaussian Process Model Predictive Control

Abstract

Due to the nonlinearity and the susceptibility to wind disturbance in actual flight, it is difficult to obtain an accurate model of the quadrotor. In this paper, a trajectory tracking control method for quadrotor based on Gaussian Process (GP) Model Predictive Control (MPC) is proposed. First, a simplified dynamic model of quadrotor is established and the unmodeled dynamic system is learned by GP models. This data-driven modeling method not only makes the modeling more accurate, but also considers the model uncertainty with covariance. Then, according to the nominal and learned GP models, model predictive controllers are separately designed for the translation and rotation subsystem of the quadrotor to track the trajectory. Finally, simulation results demonstrate the effectiveness of the proposed approach.