Quadrotor Trajectory Tracking Control using Non-Linear Model Predictive Control with ROS Implementation

摘要

This paper presents a Model Predictive Controller for trajectory tracking control of the quadrotor using the ACADO Toolkit on Matlab/Simulink. Model Predictive Control (MPC) prediction feature and ability to obtain optimal control action yields an accurate trajectory tracking performance. The controller is applied to a quadrotor system. The mathematical model was derived using Newton's and Euler's laws. Simulations for the trajectory tracking test was done for evaluating the trajectory tracking performance. Afterwards, an interface between RotorS Gazebo Simulator and Simulink was implemented using the Robot Operating System (ROS) for validation of the controller's performance. The paper presents the results of both simulations under disturbances to determine the suitability and validity of the proposed control algorithm.