Navigation of a miniaturized autonomous underwater vehicle exploring waters under ice

Enceladus and Europa, the icy moons of Saturn and Jupiter, are among the most promising candidates in the quest for extraterrestrial life. Their giant ice shells probably conceal gigantic oceans with habitable conditions. The Technologies for Rapid Ice Penetration and Subglacial Lake Exploration (TR...

Full description

Bibliographic Details
Main Author: Nitsch, Maximilian
Other Authors: Abel, Dirk, Jeinsch, Torsten
Format: Doctoral or Postdoctoral Thesis
Language:English
Published: RWTH Aachen University 2024
Subjects:
AUV
GNC
Online Access:https://publications.rwth-aachen.de/record/987969
https://publications.rwth-aachen.de/search?p=id:%22RWTH-2024-05964%22
Description
Summary:Enceladus and Europa, the icy moons of Saturn and Jupiter, are among the most promising candidates in the quest for extraterrestrial life. Their giant ice shells probably conceal gigantic oceans with habitable conditions. The Technologies for Rapid Ice Penetration and Subglacial Lake Exploration (TRIPLE) project line aims to develop key technologies for their exploration. One core component is the TRIPLE-nanoAUV - a miniaturized Autonomous Underwater Vehicle (AUV) currently under development for a demonstration at the Neumayer Station III in the Antarctic, marking a crucial step towards exploring the mysteries of these distant, potentially habitable environments. This thesis lays the foundation for the navigation filter of the nanoAUV, which is part of the Guidance, Navigation, and Control (GNC) system. Due to the absence of satellite navigation systems, underwater navigation is quite challenging, especially for small AUVs. This thesis, therefore, identifies a highly miniaturized sensor package for the navigation of the nanoAUV. Its core sensor is an Inertial Measurement Unit (IMU) whose integrated measurements are corrected by an Ultra-Short Baseline (USBL) system, a Doppler Velocity Log (DVL), a depth pressure sensor, and a magnetometer.For sensor fusion, a classic ES-EKF and a UKF are developed, validated with the help of high-fidelity sensor models and Monte Carlo simulations, and evaluated experimentally with data from a field test in the Mediterranean Sea. The results demonstrate that positioning accuracies in the decimeter range can be achieved, enabling a safe return of the nanoAUV to the docking module. The thesis particularly emphasizes the real-time capability of the navigation algorithms. Additionally, an extension of the ES-EKF framework, namely the invariant EKF (InEKF), is presented based on the theory of manifolds, which only marginally increases the computational effort. The results show that the InEKF outperforms the ES-EKF and the UKF under strongly perturbed initial conditions. The ...