ESR3 - Reduced-order models and machine learning for FOWT analysis and design

Aims: This ESR will develop reduced order models for the modelling of floating wind turbines. This will be based on enhancing low-fidelity physical models with high-fidelity data, hence retaining physical correctness while matching reference data. The resulting models are expected to extrapolate better and require less training data. New reduced models will be derived for a wind turbine undergoing wave motions. As such, ESR3 will build a bridge between high-fidelity studies (ESR1&2) and design methods (ESR4). Machine-learning will be used on the training-data to build improved closure models.

Work package: WP1 - Design

Beneficiaries: TU Delft, Siemens Gamesa Renewable Energy

Axelle Viré
Assistant Professor

Project Coordinator, expert in computational fluid dynamics and fluid-structure interactions