Repository logo
 

Energy-Optimal Control for Unmanned Underwater Vehicles in Offshore Aquaculture

Files

Thesis embargoed until 18th November 2028(26.62 MB)

Date

2025

Authors

Tun, Thein Than

Supervisor

Huang, Loulin
Singamneni, Sarat

Item type

Thesis

Degree name

Doctor of Philosophy

Journal Title

Journal ISSN

Volume Title

Publisher

Auckland University of Technology

Abstract

With sustainable land-based farming and offshore aquaculture, the New Zealand aquaculture industry is expected to reach $3 billion market size by 2035 [1]. Aligning with this trend, the New Zealand King Salmon (NZKS), a leading fish farming company, started the proof-of-concept phase of its offshore aquaculture project called Blue Endeavour Project (the first of its kind offshore salmon farm in New Zealand) in June 2025 [2]. Along with relocation to the offshore, unmanned underwater vehicles (UUVs) play a crucial role in conducting work processes such as fish net-pen visual inspection (FNVI) autonomously in collaboration or on behalf of human workforce due to the harsh working environment. However, due to the constrained operational workspace in aquaculture, the autonomous UUV must rely on its limited onboard battery capacity, without the use of an umbilical cord for power supply and data transmission. Therefore, the energy-optimality aspect of optimal control is investigated in this thesis. The findings of this research are expected to contribute to the growing adoption of autonomous UUVs in aquaculture and oceanographic research, where energy efficiency and accurate trajectory tracking in constrained operational workspaces are essential, thereby enabling fewer battery-charging cycles and extended operational duration and range. Three main control schemes, namely, Proportional Integral Derivative-based controllers (PID-based controllers), Linear Quadratic Tracking-based controllers (LQT-based controllers), and Model Predictive Control-based controllers (MPC-based controllers), are applied in a total of 13 controllers to two UUVs, namely BlueROV2 Heavy Configuration (an inspection class UUV, weighing about 12 kg) and RexROV 2 (an intervention class UUV, weighing about 1800 kg), to conduct FNVI around NZKS’s Blue Endeavour Project site. Explicit power functions (component-level power function and system-level power function) are used in the performance index (also known as cost function or objective function) of the energy-optimal versions of LQT-based controllers and MPC-based controllers. Among all the proposed controllers, particularly for a harsh ocean environment with strong underwater current disturbances, the energy-optimal MPC (EO-MPC) on RexROV 2 is the most energy-optimal controller, achieving the equivalent trajectory tracking performance of conventional MPC (CO-MPC) for the energy-demanding trajectory while fulfilling constraints across both prediction and control horizons. To access the realistic phenomenon of UUV’s performance in the field-trials, high-fidelity simulations with 0.0m/s - 0.9m/s underwater current disturbance speeds are conducted in the Robot Operating System (ROS), integrated with Gazebo Physics Engine, using the specifications of UUV’s system parameters and NZKS’s Blue Endeavour Project. Due to the unavailability of an expensive and accurate localisation system for the sea/ocean trials, a total of 13 pool experiments were conducted using the actual hardware of BlueROV2 with the vision-based state-estimation system. These experimental results are cross-validated with another 13 pool simulation results and it was found that both experimental and simulation results are, in general, coherent among themselves and FNVI simulations. Most importantly, it is worth noting that the types of UUVs, the available computing capacity on the chosen UUV, the operational constraints, and environmental factors must be taken into account in choosing a controller among the proposed energy-optimal controllers. In addition, these research findings can be used as part of the technical evidence and operational guidelines by multiple stakeholders (e.g., policymakers, infrastructure owners, and management, development, and deployment engineers), thereby enabling comprehensive and informed decision-making regarding the advantages and challenges associated with the adoption of autonomous UUVs in offshore aquaculture.

Description

Keywords

Source

DOI

Publisher's version

Rights statement

Permanent link

http://hdl.handle.net/10292/20132

Collections

Doctoral Theses