Robot Assisted Landing Process of small UAVs using Decentralized Kalman Filter
Year: 2022
Type: conference-paper
Abstract: This paper presents a new approach on robot-assisted landing of a small unmanned tilt-wing aircraft using a serial kinematic industrial robot arm to directly catch the aircraft in hover flight. The goal is to automate the landing and turnaround of the aircraft without the need for a large landing area and additional personnel. For this purpose, a base station is presented that integrates the robot arm along with other required hardware. An optical tracking system is used to acquire the aircraft pose. The catching process itself is formalized in three successive phases, each with its own set of constraints and tolerances. A direct tracking method for executing the catching process is proposed and evaluated through a real-time pose tracking setup with the robotic arm. The results indicate the necessity of a filtered setpoint pose as well as a short term prediction to overcome delays in the robot controller. Therefore, a decentralized Kalman filter is developed to obtain a robust estimation of the aircraft pose by incorporating measurements from a ground-based optical tracking system as well as measurements from the inertial measurement unit streamed directly from the aircraft. In addition, an outline on further work for robot trajectory generation is given.
Source: EUROGNC
Authors: Konnow, Vincent, Terlau, Jannes, Hildebrand, Lukas, Hartmann, Philipp
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6902951283