AI-Driven fault diagnosis for wind turbine generators (ABDIJALEBIS _U27EMPSFP1)
Key details
- Application deadline
- 31 July 2026 (midnight UK time)
- Location
- UEA
- Funding type
- Self-funded (Home students only)
- Start date
- 1 October 2026
- Mode of study
- Full-time or part-time
- Programme type
- PhD
)
Welcome to Norwich
According to the Sunday Times, this city is one of the best places to live in the UK.
Project description
Primary supervisor - Dr Abdi Jalebi Salman(opens in a new window)
Reliability is critical to the continued expansion of wind power, particularly for offshore installations, which already contribute around 10-15% of the UK’s electricity and are central to achieving net-zero targets and ensuring the resilience of national grid infrastructure. Preventive maintenance enabled by condition monitoring systems (CMSs) plays a key role in improving turbine reliability and availability while reducing operational costs and the levelised cost of energy. However, the drivetrain, comprising the gearbox, generator, and power electronics, remains a major source of failures, accounting for roughly one third of incidents and nearly half of maintenance costs in offshore turbines. Early and accurate fault detection in this subsystem is therefore essential.
Despite widespread deployment, existing CMS solutions are predominantly based on vibration measurements from accelerometers mounted on drivetrain components, and their performance remains limited. These systems often suffer from low fault detection rates, generate false alarms that reduce energy production, or fail to identify faults sufficiently early to avoid costly repairs. Moreover, they are largely ineffective at detecting electrical faults. The lack of commercially available CMS solutions based on electrical measurements is due to inherent complexity of interpreting electrical signatures of mechanical faults in real time. This creates a clear gap and opportunity for more advanced and integrated monitoring approaches.
This project aims to address the identified gap by developing hybrid fault diagnosis methods that combine analytical approaches based on drivetrain physical properties with AI-driven data analysis techniques to enhance the accuracy and effectiveness of fault detection and classification. The work will involve analytical studies, computer simulations, and finite element (FE) analysis, alongside experimental development and validation.
Entry requirements
The minimum entry requirement is 2:1 in Electrical and Electronics Engineering, Mechanical Engineering, Physics.
Funding
This is a self-funded project.
References
i)
N.K. Hosseini, H. Toshani, S. Abdi, “Enhanced Bearing Fault Detection in Induction Motors Using Projection-Based SVM”, IEEE Transactions on Industry Applications, DOI: 10.1109/TIA.2025.3536425, February 2025
ii)
A. Afshar, S. Abdi, A. Oraee, R. McMahon, “Eccentricity fault detection in brushless doubly fed machines”, IET Electric Power Application (EPA), Vol. 15, p.p. 916-930, March 2021.
iii)
N. Khadem, H. Toshani, S. Abdi, “A Projection-Based Support Vector Algorithm for Induction Motors’ Bearing Fault Detection”, IEEE International Symposium on Diagnostics for Electric Machines, Power Electronics and Drives (SDEMPED), Greece, August 2023.
iv)
M. Vazifehdan, H. Toshani, S. Abdi, “Mechanical fault detection in induction motors using data-driven Kalman filter”, IEEE International Symposium on Diagnostics for Electric Machines, Power Electronics and Drives (SDEMPED), Greece, August 2023.
v)
S. Abdi, S. Sharifzadeh, S. Amiri, “Reliability Model Development for Wind Turbine Drivetrain with Brushless Doubly Fed Induction Machine as Generator”, IEEE International Conference on Industrial Technology (ICIT), Valencia - Spain, March 2021.