Overview
This project is part of a 4 year Dual PhD degree programme between the National Tsing Hua University (NTHU) in Taiwan and the University of Liverpool in England. As Part of the NTHU-UoL Dual PhD Award students are in the unique position of being able to gain 2 PhD awards at the end of their degree from two internationally recognised world leading Universities. As well as benefiting from a rich cultural experience, Students can draw on large scale national facilities of both countries and create a worldwide network of contacts across 2 continents.
About this opportunity
Power systems are undergoing a rapid transition toward high penetrations of variable renewable energy sources such as wind and solar. This transition introduces significant challenges in maintaining power grid flexibility and balancing supply and demand in real time. Traditionally, flexibility has been provided by large, centralised generators; however, the growing deployment of smart technologies and distributed energy resources enables residential consumers to play an active role in balancing the grid and supporting a more sustainable energy system. By adjusting their energy consumption or generation at optimal times, through demand-side flexibility, households can reduce peak demand, improve grid efficiency, lower reliance on carbon-intensive generators, and promote greater renewable energy integration and decarbonisation.
However, achieving reliable residential flexibility remains difficult and complex. Human behaviour is uncertain and interactive, while comfort and privacy constraints must be respected. Current flexibility models and control algorithms often overlook these user-centric factors, treating end-users as passive system components. Moreover, existing methods primarily rely on forecasting and optimisation, where uncertainties are penalised economically or lead to network constraint violations. These limitations result in suboptimal and potentially unsafe control decisions that hinder large-scale adoption and slow the transition to sustainable energy systems.
This project proposes to address these challenges by developing a safe, coordinated, and decision-focused machine learning framework that integrates expertise from computer science and electrical engineering. The research will explore safe reinforcement learning to ensure user comfort and system reliability, use game theory to model the interactions between users, and apply graph-based learning to capture relationships among households with shared feeders, tariffs, or behavioural similarities. By embedding these techniques within realistic power system models, the project aims to produce trustworthy, human-aware, and scalable algorithms for residential participation in smart grid flexibility services, bridging the gap between technical efficiency, user acceptance, and environmental sustainability in future energy systems.
This PhD is delivered through the dual NTHU–University of Liverpool programme. The first two years will be spent at The University of Liverpool (Dr Chao Long) and the following two years at NTHU (Prof Wing-Kai Hon). This collaboration showcases strong complementary expertise and therefore offers a unique opportunity for this project which covers the power system/smart grid, computer science and AI algorithms.