Development of a formation control algorithm to coordinate multiple biomimetic AUVs in the presence of realistic environmental constraints

McColgan, Jonathan (2019) Development of a formation control algorithm to coordinate multiple biomimetic AUVs in the presence of realistic environmental constraints. PhD thesis, University of Glasgow.

Full text available as:
[img]
Preview
PDF (edited version, 3rd party copyright removed)
Download (17MB) | Preview
Printed Thesis Information: https://eleanor.lib.gla.ac.uk/record=b3342986

Abstract

Biomimetic Autonomous Underwater Vehicles (BAUVs) are a class of Uncrewed Underwater Vehicle (UUV) that mimic the propulsive and steering mechanisms of real fish. However, as with all UUVs, the range and endurance of these vehicles remains limited by the finite energy source housed on board the vehicle. Unsurprisingly, a consequence of this finite energy source is that BAUVs/UUVs are incapable of completing the large-scale oceanographic sampling missions required to drastically improve our understanding of the Earth’s oceans and its processes. To overcome this limitation, this thesis aims to investigate the feasibility of deploying a self-coordinating group of BAUVs capable of completing the aforementioned oceanic surveying missions despite the constraints of the local operating environment.
To achieve this, the work presented in this thesis can be separated into four distinct parts. The first of which is the development of a suitable mathematical model that accurately models the dynamics of the RoboSalmon BAUV designed and built at the University of Glasgow. As well as ensuring the models validity, its ability to efficiently simulate multiple vehicles simultaneously is also demonstrated.
The design and implementation of the formation control algorithm used to coordinate the vehicles is then presented. This process describes the alterations made to a biologically-inspired algorithm to ensure the required parallel line formation required for efficient oceanic sampling can be generated. Thereafter, the implementation of a realistic representation of the underwater communication channel and its debilitating effect on the algorithms ability to coordinate the vehicles as required is presented.
The thesis then describes the incorporation of two methodologies designed specifically to overcome the limitations associated with the underwater communication channel. The first of which involves the implementation of tracking/predictive functionality while the second is a consensus based algorithm that aims to reduce the algorithms reliance on the communication channel. The robustness of these two methodologies to overcoming not only the problematic communication channel but also the inclusion of additional external disturbances is then presented.
The results demonstrate that while the tracking/predictive functionality can overcome the problems associated with the communication channel, its efficiency significantly reduces when the external disturbances are taken into consideration. The consensus based methodology meanwhile generates the required formation regardless of the constraints imposed by both the communication channel and the additional external disturbances and therefore provides the more robust solution.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Additional Information: Due to copyright issues the full electronic version of this thesis is not available for viewing. An edited version (3rd party copyright removed) is available.
Keywords: Biomimetic, formation control algorithms, AUVs, robotics, mathematical modelling, underwater communication channel.
Colleges/Schools: College of Science and Engineering > School of Engineering > Aerospace Sciences
Funder's Name: Engineering and Physical Sciences Research Council (EPSRC)
Supervisor's Name: McGookin, Dr. Euan
Date of Award: 2019
Depositing User: Mr Jonathan McColgan
Unique ID: glathesis:2019-41103
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 04 Apr 2019 13:10
Last Modified: 30 Apr 2019 10:37
URI: http://theses.gla.ac.uk/id/eprint/41103
Related URLs:

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year