Session: 41-03 AI and advanced methods for wind energy II
Submission Number: 175540
Closed-Loop Ai-Driven Optimization of Wind Turbine Blades Using CFD, Neural Networks, and Experimental Validation
Wind turbine blade design is a central challenge in advancing renewable energy, as it must strike a balance between aerodynamic efficiency, noise reduction, and manufacturability. Traditional optimization approaches based solely on computational fluid dynamics (CFD) suffer from high computational expense, while purely experimental methods are slow and difficult to scale. This work introduces a closed-loop optimization framework that integrates CFD, neural network–based surrogate modeling, and iterative experimental validation. The process begins with CFD simulations across a representative design space to generate aerodynamic and aero-acoustic data, which are then used to train neural networks capable of predicting lift, drag, and noise responses with reasonable accuracy at reduced computational cost. Candidate geometries proposed by the surrogate are fabricated using additive manufacturing and tested in a controlled wind tunnel environment, with experimental results fed back into the loop to refine the model. The framework reduced the CFD workload by 40–60% compared to brute-force approaches and produced blade designs with 5–10% improvements in the lift-to-drag ratio. Experimental measurements also indicated consistent broadband noise reductions under design operating conditions, though benefits diminished at off-design wind speeds. Overall, the results demonstrate that combining simulation, AI-based prediction, and experimental feedback offers a practical and scalable pathway for accelerating wind turbine blade development and improving renewable energy system performance.
Presenting Author: Ryo Amano University of Wisconsin-Milwaukee
Presenting Author Biography: Professor
Authors:
Abdallah Benelmadjat University of Wisconsin-MilwaukeeSaif Al Hamad University of Wisconsin-Milwaukee
Omar Shaker University of Wisconsin-Milwaukee
Ryo Amano University of Wisconsin-Milwaukee
Closed-Loop Ai-Driven Optimization of Wind Turbine Blades Using CFD, Neural Networks, and Experimental Validation
Paper Type
Technical Paper Publication