Model Predictive Control with Compensation of Nonlinearities for Quadcopter Control
Nonlinear Model Predictive Control, Robotics, Qhadcopter, UAV.
This study analises the application of Model Predictive Control (MPC) with the proposed Nonlinear Compensation Method in order to control a quadcopter. This method is based on the use of the system nonlinear model to predict and insert nonlinearities in the predictive control formulation by means of an affine structure. With respect to the control of a quadcopter, the proposed controller is compared, using numerical simulations, to the traditional linear predictive controller. After that, experiments were made with a real quadcopter mounted in a test bench. As a result, the use of the proposed method brought a considerable perfomance improvement when the system state were farther apart from the operation point in which it was linearized.