Speaker
Description
This study introduces an innovative methodology for enhancing the performance of wind turbines by optimizing the distribution of chord length and twist angle. The approach integrates a multi-objective genetic algorithm (MOGA) with a screening method, iteratively refining design candidates through a series of optimization cycles. These candidates are subsequently evaluated within a Design of Experiments (DOE) framework. The optimization process employs a surrogate model built using the Kriging method to construct a response surface, facilitating accurate performance predictions. The computational fluid dynamics (CFD) solver, based on Reynolds-Averaged Navier-Stokes (RANS) equations, is leveraged to evaluate aerodynamic performance within the optimization process. The proposed methodology is rigorously compared with standalone MOGA and screening techniques to assess its efficacy. The results demonstrate a notable 15% performance improvement over the baseline rotor design, along with a broader and more reliable design space compared to MOGA and screening methods individually. This hybrid approach not only enhances optimization efficiency but also contributes to a more robust exploration of design alternatives.
