PhD Studentship - Machine Learning for Power System Dynamics

How to apply: uom.link/pgr-apply-fap

This 3.5 year PhD is fully funded; the tuition fees are paid and you will receive an annual tax free stipend set at the UKRI rare (£19,237 for 2024/25). The funding is linked to UK Research and Innovation Future Leaders Fellowship (gtr.ukri.org/projects?ref=MR%2FY00390X%2F1). The funding is for home students.

We are looking for a PhD candidate with a strong mathematical background, curiosity and interest in applying machine learning to power systems.

You will be part of a strong power systems group (www.eee.manchester.ac.uk/research/expertise/energy-networks) at the University of Manchester, working with a team led by Dr. Panagiotis Papadopoulos built around a UK Research and Innovation Future Leaders Fellowship focusing on “Addressing the complexity in future power system dynamic behaviour” (gtr.ukri.org/projects?ref=MR%2FS034420%2F1 and gtr.ukri.org/projects?ref=MR%2FY00390X%2F1)

The project will be supervised by Dr. Panagiotis Papadopoulos and Prof. Jovica Milanovic.

Project background and details:

Power systems are going through unprecedented changes, mainly driven by the need for decarbonisation. This leads to the connection of several new types of devices, including renewable generation, electric vehicles, HVDC interconnectors, etc. These devices are mostly power electronic interfaced introducing new types of dynamic phenomena and the need for more detailed models, increasing complexity. In addition, intermittent behaviour of renewable generation but also social aspects and market structures related to how we use electricity, increase uncertainty.

Power systems are inherently nonlinear dynamical systems, requiring large computational effort to assess and study system stability. This is becoming even more challenging under increasing complexity requiring detailed dynamical models and with new dynamic phenomena arising (e.g. new types of oscillatory phenomena). In addition, a much larger number of scenarios need to be investigated due to increasing uncertainty in power system operation and lack of knowledge on where worst-case scenarios lie.

This PhD project will investigate the use of state-of-the-art machine learning techniques for calculating the stability boundary of complex nonlinear dynamical systems. Going beyond the notion that machine learning models are just powerful black box predictors, the project will also consider methods that can take into account physics and focus on aspects related to trustworthiness.

Applicants should have, or expect to achieve, a first honours degree or a master’s (or international equivalent) in a relevant science or engineering related discipline.

Please contact the supervisor, Dr. Panagiotis Papadopoulos (panagiotis.papadopoulos@manchester.ac.uk), before you apply. Please include a copy of your CV with details of your current level of study, academic background and any relevant experience and 1-page cover letter describing your motivation and why you are appropriate to study this PhD project.

Advert information