Fast aerodynamic analysis method for three-dimensional morphing wings based on deep learning

Abstract

Morphing wings have garnered widespread attention due to their superior aerodynamic efficiency. However, in the design process, accurately and efficiently obtaining the three-dimensional flow field of morphing wings remains a challenging issue. This paper proposes a Deep Learning-based method for predicting the flow field of a Biomimetic Morphing Wings to address this problem. Firstly, a Coordinate Transformation Mechanism is established for the studied Biomimetic Morphing Wing to ensure the consistency of grid point coordinates between different wing shapes. Secondly, a two-level Flow Field Prediction Model is constructed, consisting of grid point prediction level and physical quantity continuity adjustment level. Using this method, the flow field of the Biomimetic Morphing Wing was predicted, and the predication result were similar to those of numerical simulations. This indicates that the proposed method maintains high prediction accuracy while reducing computation time, thereby enhancing the analysis efficiency of the morphing wing's flow field.