Hybrid Physics-Informed Machine Learning for Multi-Vector Energy System Optimisation

The PhD will be based in the School of Electrical and Mechanical Engineering and will be supervised by Dr Shamsul Masum, Dr Anton Hettiarachchige Don and Dr Edward Smart.

The UK’s shift to net-zero energy requires optimising complex multi-vector networks—electricity, heat, gas, and hydrogen—under high renewable penetration. Traditional physics-based models are accurate but slow, while purely data-driven methods are fast but may violate physical laws. This project develops hybrid physics-informed machine learning models that combine physical principles with neural networks to ensure constraint satisfaction, robustness, and computational efficiency. Models will be validated under diverse conditions, producing scalable, interpretable tools to enhance optimisation and decision-making for future decentralised energy systems.

The work on this project will:

• Comprehensive literature review on physics-informed machine learning, multi-vector energy modelling, and hybrid optimisation techniques.
• Development of constrained neural network architectures incorporating energy system fundamentals.
• Implementation of hybrid models in using deep learning frameworks and optimisation libraries.
• Comparative benchmarking against conventional approaches under diverse operational scenarios.
• Validation studies assessing model interpretability, constraint satisfaction, and generalisation performance.

Project description

The increasing complexity of modern energy systems, integrating electricity, heat, gas, and hydrogen vectors alongside intermittent renewables, demands advanced modelling and control approaches that can operate in real-time whilst maintaining physical plausibility. Whilst pure data-driven artificial intelligence methods offer computational speed and pattern recognition capabilities, they lack interpretability and can violate fundamental physical laws. Conversely, traditional physics-based models provide rigorous guarantees but struggle with computational tractability and parameter uncertainty in large-scale systems. This research addresses the critical need for hybrid approaches that combine the strengths of both paradigms.

Multi-vector energy systems present unique challenges: tight coupling between energy carriers, diverse temporal dynamics spanning milliseconds to seasons, nonlinear component behaviours, and operational constraints spanning technical, economic, and regulatory domains. Current optimisation frameworks either rely on computationally expensive detailed simulation or employ oversimplified data-driven surrogates that may produce physically infeasible solutions. Neither approach adequately addresses the real-time decision-making requirements of future decentralised energy infrastructure.

This research proposes to develop novel hybrid modelling architectures that integrate physical principles with machine learning capabilities. The challenge lies in balancing model accuracy, computational efficiency, constraint satisfaction, and interpretability, all essential requirements for deployment in safety-critical energy systems. The approach must handle uncertainties in renewable generation, demand patterns, and system parameters whilst guaranteeing physically admissible solutions.

Validation will employ operational data from energy networks experiencing high renewable penetration and complex multi-vector interactions. The research must demonstrate applicability across diverse operational scenarios, including normal operation, stress conditions, and edge cases not represented in training data. Particular emphasis will be placed on model interpretability and explainability to enable regulatory acceptance and operator confidence.

Research outcomes will advance theoretical understanding of physics-informed learning for energy applications and provide practical frameworks for managing increasingly complex energy infrastructure in the UK's transition to net-zero.

General admissions criteria

You'll need a good first degree from an internationally recognised university (minimum upper second class or equivalent, depending on your chosen course) or a Master’s degree in a related area. In exceptional cases, we may consider equivalent professional experience and/or Qualifications.

English language proficiency at a minimum of IELTS band 6.5 with no component score below 6.0.

International students will require a study visa from UKVI to pursue the degree in the UK. If the research is in a sensitive or technological subject, the student may also need to secure an Academic Technology Approval Scheme (ATAS) certificate from the UK Foreign Office.

How to Apply

We’d encourage you to contact Dr Shamsul Masum (shamsul.masum@port.ac.uk) to discuss your interest before you apply, quoting the project code.

When you are ready to apply, please follow the 'Apply now' link on the Electronic Engineering PhD subject area page and select the link for the relevant intake. Make sure you submit a personal statement, proof of your degrees and grades, details of two referees, proof of your English language proficiency and an up-to-date CV. Our ‘How to Apply’ page offers further guidance on the PhD application process.

When applying please quote project code: SEM10450526