디지털 트윈 구축을 위한 메쉬 기반 조립품 모델 CAD 최적화에서 군집화의 적용

Abstract

In the engineering field, three-dimensional (3D) computer-aided design (CAD) models are employed toconstruct a digital twin, and controlling the level of details of 3D CAD models is necessary to visualizelarge-scale 3D CAD models efficiently. In general, mesh models are most frequently used for visualizing 3Dmodels. However, studies on simplification of mesh-based 3D CAD assembly models are scarce. Therefore,this study was focused on assessing the feasibility of applying clustering algorithms to simplify mesh-based3D CAD assembly models. We first proposed a method that utilizes region growing to cluster parts in anassembly model. After clustering, mesh optimization was performed by applying the ant colony optimizationalgorithm to determine the optimal simplification ratio for each cluster. We conducted experiments using fivetest models, which can be used in the engineering field. In conclusion, the proposed method outperformed theexisting methods both quantitatively and qualitatively.