Professor Mahmoud Mostafavi, University of Bristol

UK is world-leading in fusion science and engineering and is set to build the first fusion power plant by 2040. This is an ambitious time-line which challenges the conventional methods of engineering design and fabrication. Therefore, to achieve this goal, innovative step change is needed in the way we design a power plant. Recent advances in computational power and the expansion of application of machine learning methods in engineering has enabled us to carry out significant part of design and its optimisation in-silico. This project aims to develop physics-based simulations for integrity of critical components in a fusion power plant (e.g. a diverter) and use machine learning algorithms methods (e.g. Gaussian Process) to build fast running surrogate models of physics-based simulations to enable optimisation of such components.