Machine Learning and Optimisation-Based Intelligent Substation Design in Cyber-Physical Power Systems

About the Project

Research area and project description:

This is an exciting PhD project to be carried out in close collaboration with a UK electrical network operator and an industrial co-supervisor. Power grids are undergoing rapid digitalisation with the ambition to achieve net-zero goals set by several nations. In response, grid operators are designing next-generation intelligent substations (ISS), which are the technical foundation for an automated and adaptable electricity network. They minimise asset unavailability and downtime, enabling real-time adaptive network control. They produce a resilient, long-term cost-benefit case for modernising and protecting our grid.

The project will complement the ISS initiative and model data-driven grid optimisation services that ISSs can enable. Firstly, the project will develop machine learning based solutions for predictive grid analytics (such as grid congestion forecast, asset monitoring, etc.). Based on these results, the project will develop optimisation algorithms to dynamically reconfigure the substation/distribution network settings to enhance the system efficiency. The optimisation algorithms will incorporate the uncertainties associated with the machine learning based forecasts. Furthermore, since optimisation involves data exchange between different entities in the grid, data exchange cybersecurity is an important concern. The PhD project will explore solutions to implement grid optimisation in a cyber-secure manner, such as enabling encryption services, and quantify the associated trade-offs. The results will be validated using datasets provided by a UK-based electrical network operator on real-world ISS deployment and an advanced sandbox (testing) environment to validate the project’s findings in emulated settings. The results will quantify the financial and energy gains due to the ISS deployment.

Scholarship:

The award will cover the UK tuition fee level, plus a tax-free stipend, currently £20,780, paid at the prevailing UKRI rate for 3.5 years of full-time study.

Eligibility:

The candidate should have a good 2.1 Bachelor's, or Master's degree in Electrical Engineering, Computer Science, Physical Sciences or equivalent. Experience in one of the following topics will be a plus:

• Application of machine learning to power grids/cyber-physical systems
• Knowledge of working with MATLAB/Python
• Power grid optimisation/control
• Experience working with power grid simulators such as Digsilent Power Factory or PSCAD.

How to apply:

Candidates should submit an expression of interest by sending a CV and supporting statement outlining their skills and interests in this research area to https://www.warwick.ac.uk/engineeringscholarships/sl/appcv. If this initial application is successful, we will invite you to formally apply. All candidates must fulfil the University of Warwick entry criteria and obtain an unconditional offer before commencing enrolment.