Posters

Abstract

Offshore wind turbines operate in harsh and variable environments where component failures can lead to costly downtime and repairs. Developing effective fault detection and isolation (FDI) strategies is therefore critical to ensure safety, reliability, and cost-efficiency. This study evaluates state-of-the-art FDI approaches targeting four key subsystems: mooring lines, dynamic power cables, gearboxes, and electric generators. A multi-domain framework was established to integrate mechanical, electrical, and structural health monitoring signals. Mooring line integrity was assessed using tension measurements and vibration signatures, with model-based estimators applied to detect early degradation. Dynamic cables were monitored through electrical resistance changes and partial discharge detection. Gearbox faults were identified via advanced vibration analysis and machine learning classifiers, while electric generator anomalies were isolated using current and flux signal analysis. Results show that hybrid approaches—combining physics-based models with data-driven learning—achieve higher sensitivity to incipient failures while reducing false alarms. The findings highlight the trade-off between detection accuracy and computational cost, emphasizing the need for adaptive methods that evolve with changing operating conditions. The study demonstrates how tailored FDI strategies across subsystems can extend turbine lifetime, reduce unplanned maintenance, and improve levelized cost of energy. These results support the growing adoption of predictive maintenance frameworks in offshore wind, moving toward resilient and cost-competitive energy production.