UAV Push Recovery Operation by Symmetrical Control and Estimation in Receding Horizon

Abstract

This paper presents an unmanned aerial vehicle (UAV) push recovery operation using model predictive control (MPC) and moving horizon estimation (MHE) in a symmetric manner. This utilization is motivated by the active use of UAVs, particularly for the contact based inspection of the surrounding's ceilings. To enable a physical interaction operation by an optimization-based algorithm, a primal-dual quadratic programming (QP) solver is structured for the MPC and MHE. The designed system consists of (a) an interaction model to be implemented both on the control and the estimation; (b) an integral action in the predictive controller; (c) a disturbance estimation by MHE to update the MPC. Consequently, the nominal MPC, the integral action in MPC, and the disturbance observer based MPC are compared for a UAV push recovery operation. The numerical investigations demonstrate the applicability of the proposed approach.