On the Role of Atmospheric Turbulence and Stability on Machine Learning Predictions of Wind Speed for Optimal Control of Renewable Microgrids

Effective and reliable control of wind turbines during the operation of microgrids requires accurate wind speed predictions. This study introduces an innovative approach integrating atmospheric turbulence and stability features into machine and deep learning models. Employing these features significantly improves traditional on-site wind speed prediction by employing high-frequency sensor data, leading to advancements in both primary and tertiary renewable microgrid controls. The performance metrics reveal the significant influence of incorporating turbulent fluxes and stability features into predictive models, leading to more efficient and sustainable operation of wind turbines. This research emphasizes the impactful role of machine learning algorithms, informed by high-frequency data, to contribute significantly towards optimizing renewable energy technologies and microgrid operations.