Delft University of Technology has two vacancies for PhD positions within the interdepartmental research project Model-Predictive Railway Traffic Management of the Department of Transport and Planning (Faculty CEG) and the Delft Centre for Systems and Control (Faculty 3mE), funded by the Dutch technology foundation STW.
In this project the modelling and theory of switching max-plus linear systems will be developed with a focus on computational aspects and embedding in a real-time model predictive control (MPC) framework.
The research contains three main ingredients: (1) Online parameter estimation for up-to-date train running time estimates based on train position messages; (2) Developing a real-time railway traffic prediction model based on max-plus algebra; and (3) Developing an MPC framework to keep an up-to-date timetable with respect to the current state of infrastructure, resources, and train delays, and optimize traffic control decisions in case of disruptions from a network-wide perspective. The core of the MPC approach is a discrete event dynamic system formulated as a switching max-plus linear system, which extends existing models in max-plus algebra by a switching mechanism to represent alternative control decisions or circumstances.
Candidates must have an MSc degree in mathematics, systems and control, computer science or a related field, and have preferably a strong interest in railway operations, optimization, and applied statistics. For more details, see the vacancy at TRAIL research school or contact Dr. R.M.P. Goverde,