Abstract: In the current wing design phase, acquiring extensive aerodynamic performance data of wings incurs substantial costs. To address this challenge, there is an urgent need to develop an efficient and relatively accurate method for obtaining wing aerodynamic performance data. The research focused on a low-speed double-trapezoid wing, establishing a wing dataset under subsonic flight conditions using OpenVSP and CFD simulations. Based on this dataset, a PointNet-based aerodynamic performance prediction network (PointNet-AP) was developed. The PointNet-AP network took wing point cloud data, conventional planar geometric features, and flight conditions as inputs to predict the corresponding lift coefficient (C_L), drag coefficient (C_D), and pitching moment coefficient (C_m). Experimental results demonstrated that the PointNet-AP model with transfer learning achieved satisfactory predictive performance, yielding mean relative errors of 2.88%, 4.75%, and 6.74% for C_L, C_D, and C_m predictions, respectively.In the current stage of wing design, obtaining large amounts of aerodynamic performance data is costly. This study focuses on a low-speed, double-delta wing and constructs a wing dataset for subsonic flight conditions using OpenVSP and CFD. Based on this dataset, a PointNet-based aerodynamic performance prediction network (PointNet-AP) is designed. The network predicts the lift coefficient (C_L), drag coefficient (C_D), and pitching moment coefficient (C_m) for the wing under given flight conditions, using wing point cloud data, traditional geometric features, and flight conditions as inputs. Experimental results show that the PointNet-AP model, after transfer learning, offers good prediction accuracy, with average relative errors of 2.88%, 4.75%, and 6.74% for C_L, C_D, and C_m, respectively. The prediction network proposed in this paper can lay a technical foundation for the development of large-scale aerodynamic performance prediction models by providing capabilities for data feature extraction and model transfer.