Research on the optimization of wind turbine blades using Kriging models and multi-objective genetic algorithm (MOGA)
Wu Luo, Yan Wang · Journal of Physics Conference Series · 2025
Abstract To address the critical challenges of lightweight design and cost-effective manufacturing for vertical-axis wind turbines (VAWTs), a novel structural optimization framework integrating surrogate modeling and multi-objective optimization was developed. This framework simulates operational conditions at 15 m/s wind speed and 14.3 rad/s rotor speed. By combining a Kriging response surface model with the multi-objective genetic algorithm (MOGA), a multidisciplinary optimization model was established, aiming to minimize blade mass and total deformation while constraining equivalent stress to ≤2.7516 MPa. Forty-five initial sample datasets were generated using Latin Hypercube Sampling (LHS). Global sensitivity analysis identified six core design variables, including blade length and width. Optimization results demonstrated: blade mass decreased from 1.7121 kg to 1.25 kg (26.99% reduction), equivalent stress dropped from 1.1775 MPa to 0.61928 MPa (47.41% reduction), total deformation reduced from 0.0081915 mm to 0.0027028 mm (67% reduction). Innovative contributions include: development of an aerodynamic-structural coupled multi-objective collaborative optimization framework; establishment of a parameter identification mechanism integrating experimental design and evolutionary algorithms; achievement of synergistic optimization between lightweight design and mechanical performance, providing an effective engineering solution for high-performance, low-cost VAWT design.