A Method for Skin Model Shape Generation Based on Multi-Surrogate Model Fusion and Point Clouds

Recently, skin model shapes have shown advantages and promising prospects in the research fields of tolerance analysis and digital twin. However, due to the inevitable errors during the manufacturing process, the actual surface has complex geometric features. How to establish a real surface model containing geometric errors has always been a concern for both industry and academia. In this paper, we propose an ensemble-based surrogate modeling approach to generate high-fidelity skin model shapes from sparse point cloud data. Individual surrogate models are fused in a linear weighted manner to take advantage of the different models. The verification results demonstrate that the ensemble surrogate model achieves higher accuracy than individual surrogate models. The proposed method enhances accuracy while minimizing reliance on measurement data, offering a superior foundational model for surface-related research applications.