A sub-structuring approach to overcome model limitations for input-state estimation of offshore wind turbines

Journal article

The Augmented Kalman Filter (AKF) has been applied previously for input-state estimation of offshore wind turbines (OWT). However, the accuracy of the estimated results depend on the chosen model, for which various complexities exist, making this a challenging task. Two of which are the lack of information required to model the Rotor-Nacelle Assembly (RNA), and the high uncertainty associated with the soil–structure-interaction (SSI). Therefore, the primary focus of this work is to avoid these limitations by considering a suitable substructure which eliminates the need to model the RNA and the SSI, thus significantly reducing uncertainties. The substructure is obtained by ‘cutting’ the OWT at the top of the tower and at the ground level. To define the model, the resulting substructure then only requires geometries and material properties for the monopile and tower; information which is often known with greater certainty. A numerical case study is presented to investigate the accuracy of the proposed approach for input-state estimation of a 15 MW OWT. A series of commonly used setups involving accelerometers and inclinometers are used and the effects on the predicted fatigue life of the structure are discussed. Additionally, a simple approximation of the wave loading is considered to estimate and account for its contribution to the dynamics of the substructure. The proposed approach is shown to be an effective solution for input-state estimation of OWTs when the RNA or SSI are unknown or associated with significant uncertainty.

Authors: Simpson, HA; Chatzi, EN; Chatzis, MN

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Elsevier
• Journal: Journal of Sound and Vibration
• Volume: 612
• Article number: 119153
• Publication date: 2025-05-05
• Acceptance date: 2025-04-23
• DOI: 10.1016/j.jsv.2025.119153
• EISSN: 1095-8568
• ISSN: 0022-460X
• Language: English
• Keywords: virtual sensing; input-state estimation; offshore wind turbines; modelling challenges; sub-structuring