A Collaboratively Iterative Sequential Approximate Optimization Method for Aerodynamic Optimization

Abstract

The sequential approximate optimization (SAO) method is widely adopted in aircraft optimization design, whose further improvement in efficacy and efficiency has encountered bottlenecks. In this paper, a collaboratively iterative sequential approximate optimization (CISAO) method with field metamodel (FM) accelerating computational fluid dynamics (CFD) is proposed for aerodynamic optimization problems to enhance the coupling between metamodels and CFD high-fidelity models (HFMs). In this method, the field metamodel is used to predict the flow field distributions with respect to the design variables to accelerate the numerical solution of CFD HFMs. Two benchmark aerodynamic optimization cases are performed to validate the proposed method. The time costs for single CFD HFMs in the two test cases are shortened by 37.76% and 12.11%, respectively, and the total time costs of the two test cases are shrunk by 64.11% and 44.47% compared with the conventional SAO method. The optimization results indicate that the proposed method significantly reduces the time costs and improves the optimization efficiency. The proposed method shows a promising prospect in aircraft optimization design.