Abstract
This study introduces a machine-learned actuator Line model (ML-ALM) for hydrokinetic turbines to enhance computational fluid dynamics (CFD) simulations of turbine wakes and performance. By integrating machine learning with the actuator line model (ALM), the ML-ALM seeks to accurately replicate the complex wake interactions and turbine performances observed in high-fidelity blade-resolved models while maintaining computational efficiency. Utilizing OpenFOAM simulations and data from blade-resolved models, the ML-ALM is trained to predict the forces exerted by the turbine blades under various flow conditions. This novel approach is validated against experimental data and traditional CFD models, demonstrating its potential to improve energy extraction efficiency and optimize turbine array configurations. The ML-ALM represents a significant advancement in hydrokinetic turbine modeling, balancing high-fidelity models’ accuracy and lower-fidelity models’ efficiency. This study not only contributes to the field of renewable energy by enhancing turbine simulation techniques but also showcases the applicability of machine learning in complex engineering systems.
