Research at Sheffield has been developing models of the railway network from a range of perspectives including smart-grid energy consumption, passenger satisfaction, and integration of rail with wider city and regional transport networks.
A focus of this work is the application of optimisation techniques (e.g. evolutionary algorithms, or Bayesian techniques) to identify high performing train and infrastructure configurations, either for existing systems or to build investment cases for new equipment.
This research will focus on combining aspects already modelled to allow multi-objective assessment of (for example) the trade-offs between energy use and passenger satisfaction. Activities may include passenger movement simulation, electrical power system modelling or developing autonomous train driving through deep reinforcement learning. Computing demands can grow rapidly with such models, so a significant aspect of the research is in formulating the problem in a tractable form, and application of parallel computing (GPUs) to speed solution within the optimisation process.
This is a self-funded project.
1st or 2:1 degree in Engineering, Materials Science, Physics, Chemistry, Applied Mathematics, or other Relevant Discipline.