Optimal Control in Stochastic Hydrodynamic Models: Rowing Across the Indian Ocean

In 2023 two rower teams will attempt a crossing of the Indian Ocean. My task will be to incorporate stochasticity into existing deterministic hydrodynamic models to allow for better recommendations for the rower teams. Upon completion of this, I will use numerical stochastic optimal control methods to optimise the path the rowers should take. The main contribution will be the development of the stochastic control system as almost all hydrodynamic models in use today do not account for uncertainty. Therefore, most optimal control methods have been developed with assumptions that we can perfectly model the underlying hydrodynamics (which is false).

Michael Nefiodovas

University of Western Australia

Michael Nefiodovas is a BPhil mathematics student and research assistant at the University of Western Australia, graduating at the end of 2023. Michael’s area of interest is in optimisation and reinforcement learning, in particular, applying optimal control and optimal decision making to real world processes. Michael has applied this theory in the context of optimal staffing and demand forecasting to save time and money in a university context. Michael’s research project will be focused on solving a stochastic shortest path optimisation problem to help in the world’s first rowing attempt from Australia to mainland Africa across the Indian Ocean.

In his downtime, you might find Michael playing netball with his team, enjoying time with his friends or listening to podcasts! Michael is always happy to chat, so reach out:).

You may be interested in

Jacob Vandenberg

Jacob Vandenberg

Performance of Multiscale Finite Element Methods
Adam Thompson

Adam Thompson

Bach Flow on Non-Product Homogeneous 4-Manifolds
Adam Bilchouris

Adam Bilchouris

Hybrid Non-Linear Statistical Methods with Applications to Modelling House Prices
Mark Youssef

Mark Youssef

Bayesian Estimation of Stationary Time Series Models with Exogenous Inputs
Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text.