Mobile robotic platforms such as USVs (Unmanned Surface Vessels) and UUVs (Unmanned Underwater Vehicles) used in offshore or open ocean aquaculture, rely on an on-board power supply (e.g., battery). This project aims to develop energy-optimal control schemes to allow a robotic platform to achieve its operation objectives with minimum energy consumption. The developed control schemes will not only extend the operation time and the working range of the robotic platform, but also reduce the size of the power supply, leaving room to enhance the payload capacity of the platform.

Almost all the man-made or natural phenomenon cannot be explicitly modelled as simple linear systems. Non-linearity exists in almost all the systems, but not all the parameters associated with nonlinearity are experimentally measurable and thus, not all the non-linear analytical models are practically applicable due to uncertainties, noises and other confounding factors which are not accurately reflected or ignored in the models. Specifically, in this proposed project, both the robotic platform and the operating environment (hydrodynamic: fluid system) are highly non-linear and high dimensional complex system. Thus, in addition to the conventional method of modelling the systems with closed form mathematical expressions derived from first principles in engineering, data-driven approach based on sensor fusion and data analytics is the trend to model and analyse complex systems. Upon building up those models, an optimal scheme developed from data driven approach and non-linear optimal control theory is proposed to control mobile robotic platforms in offshore aquaculture with minimal energy. Personally, I believe the research impact (outcome of this project) and the experience I would gain throughout my doctoral study will be highly relevant to what the research community (both industry and academics) will explore as marine robotics is an active research field to deep dive into the unexplored ocean.