Power System Vulnerability Detection and Stability Enhancement via Distributed Machine Learning

This international research collaboration, led by the University of Hertfordshire (UK) in partnership with Yildiz Technical University (Turkey), addresses a pressing challenge in the global transition to clean, sustainable energy: maintaining power grid stability as the share of variable renewable sources like wind and solar increases. In 2024, the UK witnessed record-high energy demand. With growing reliance on renewables, the grid faces heightened risks due to the intermittent and unpredictable nature of these energy sources.

While renewable energy is essential for reducing greenhouse gas emissions and achieving net-zero goals, it can cause sudden changes in electricity supply that threaten power system stability. These fluctuations can lead to voltage instability, reduced reliability, and, in extreme cases, system collapse. To address this, the project aims to develop intelligent, cost-effective tools to detect weak points in the grid and reduce instability risk. These tools will use advanced artificial intelligence, machine learning, and data analytics for a smarter, more adaptable energy infrastructure.
The research will focus on three key innovation areas:
**1- Data Monitoring Platform:** A high-performance platform will be developed to collect and analyse real-time data from energy systems. This platform will handle data from diverse sources, providing an accurate and detailed view of the grid's operational status, and potential vulnerabilities.
**2- Automated Feature Engineering:** AI techniques will be used to extract meaningful insights from data without relying on traditional physical models. This approach allows for more flexible and accurate monitoring of power system stability, as AI algorithms learn from data patterns to identify early warning signs of faults or failures.
**3- Distributed Machine Learning:** To handle the vast amount of data efficiently, the project will implement distributed machine learning across multiple computing systems. This avoids bottlenecks associated with centralised data processing and ensures high-speed performance, scalability, and efficient bandwidth usage.
At the University of Hertfordshire, Dr. Shady will serve as the Lead Principal Investigator, focusing on modelling the complex dynamics of power grids and overseeing the development of stability monitoring and mitigation tools. He will also coordinate all aspects of the UK team’s research. Dr. Amira and Dr. Mostafa will be responsible for designing and analysing grid-connected renewable systems to assess their stability under various operational scenarios. Dr. Olu will contribute expertise in machine learning and data analytics, supporting the development of intelligent monitoring platforms. At Yildiz Technical University, Dr. Nader will lead the Turkish team, bringing his expertise in renewable energy and energy storage systems to ensure proper design and implementation of the proposed solutions. Dr. Enes will support the creation of a test bed for real-time hardware-in-the-loop evaluations, helping validate system performance under realistic conditions. Dr. Faiths will also contribute significant knowledge in machine learning and data analytics for the project’s distributed platform development.

Beyond academia, the project will enhance grid reliability, help energy providers prevent blackouts, and support a smoother transition to net-zero. It will also promote innovation, workforce development, and growth in the digital energy sector—advancing secure, sustainable, and resilient energy systems in the UK and Turkey.

Stability Enhancement