The maturation of artificial intelligence (AI) and machine learning (ML) has transformed the process of data-driven science, leading to new fundamental insights and improved predictive models that can address many outstanding challenges in wind energy. In this emerging paradigm, AI/ML techniques provide a new pathway to obtaining computationally efficient surrogate models that encode the key insights from high fidelity models or experimental campaigns into design-oriented tools.
In this webinar, we highlight several successful efforts at the National Renewable Energy Laboratory (NREL) to apply AI/ML across wind energy science.
First, we demonstrate a new tool to study the impact of different climate scenarios on future wind resources using machine learning super resolution techniques. Second, we demonstrate how neural network surrogate models can enable rapid national-scale assessments of wind technology innovation like wake steering or large-scale impacts of atmospheric stability. Finally, we show how AI/ML approaches are revolutionizing turbine blade design by better representing unsteady aerodynamics in design tools and enabling direct inverse design of 3D blades with invertible neural networks. We conclude with a discussion of future research directions at the intersection of wind energy science and AI/ML.