Multi-Objective Aerodynamic Design Optimization Method Based On Variable-Fidelity Surrogate Model

Variable-fidelity modeling methods show significant promise in multi-objective aerodynamic optimization by incorporating substantial low-fidelity data to construct cost-effective, high-accuracy surrogate models. In this paper, we propose an efficient multi-objective optimization algorithm based on the hierarchical kriging (HK) model. By introducing the HK model into the multi-objective optimization algorithm, we significantly enhance optimization efficiency and accuracy. The developed algorithm is linked with Euler/Navier-Stokes solvers, geometric parameterization, automatic mesh generation, and other advanced technologies to develop a comprehensive multi-fidelity model-based multi-objective aerodynamic design optimization method. To validate the proposed method, we conduct extensive tests with a series of multi-objective test functions against traditional multi-objective evolutionary algorithms. Results conforms the method's validity and efficiency, showing superior performance in various test cases. Furthermore, we apply the method to the aerodynamic shape optimization of a wide-speed-range airfoil. This engineering application showcases the developed method's practical benefits, showing its ability to significantly enhance the aerodynamic performance of airfoils across a broad speed range while operating under limited computational resources. The method's capability to optimize aerodynamic shapes in complex and variable operating conditions underscores its substantial engineering application potential.