Robust control of an actively controlled drogue for autonomous aerial docking

Abstract

Automating the aerial docking procedure of two aircraft using the probe-and-drogue method can enable fully autonomous docking maneuvers for various applications in the future. For this purpose, a simplified system model of an actively controlled drogue was derived by combining the aerodynamic forces of control surfaces and the drogue body through superposition. Wind tunnel tests and computational fluid dynamics simulations were conducted to identify the aerodynamic coefficients, which were then used for model-based controller design. Robust control strategies such as sliding mode control, super twisting control, and PID control, which served as a baseline, were implemented and tested in a comprehensive probe-and-drogue simulation under various external disturbances. A super twisting disturbance observer was added to enhance the controllers' performance. Furthermore, all tested control architectures were modified by adding incremental nonlinear dynamic inversion to reduce perturbation rejecting controller gains while preserving control performance. The derived system model and the implemented controllers were shown to be effective for this control problem. Specifically, the application of incremental nonlinear dynamic inversion can lead to reduced control input variation without compromising control performance.