基于多层次表示的三维同步建模方法分析-analysis of 3d synchronous modeling method based on multi-level representation.docx

基于多层次表示的三维同步建模方法分析-analysis of 3d synchronous modeling method based on multi-level representation.docx

基于多层次表示的三维同步建模方法分析-analysis of 3d synchronous modeling method based on multi-level representation

针对同步建模中 CAD 模型约束驱动所面临的欠约束问题,提出基于虚几何的约束驱动方法,从 MLR 中搜索与驱动对象相关的形状与关系特征,通过这些特征到约束对 象及约束的映射,借助虚元素使约束方向与约束对象定义方向(如法矢)相一致,以 建立可求解的初始约束有向图;以此为基础,自驱动对象开始,沿约束传播方向依次 对各结点进行自由度分析以定位欠约束结点,以最小几何拓扑变更为准则对所需约束 进行几何推理,自适应地为欠约束结点增补相应虚几何约束,并通过混合约束求解, 使驱动结果更为符合设计意图要求。基于上述研究,开发三维同步建模原型系统 InteModel,实现产品 CAD 模型的快 速构建、特征提取与约束驱动,以支持产品形状的重用与再设计。最后通过工程设计 应用实例,验证了本文研究成果的正确性和有效性。关键词:三维同步建模,多层次表示,特征提取,约束驱动AbstractThe efficiency of product design can be improved by reusing CAD models. 3D synchronous technology is an important approach for model reuse and rapid design, which supports the feature extraction and constraint driven with satisfying design intent. The design intent has become diversified since the shape of product is more and more complicated. Therefore, some improved methods of feature extraction and constraint driven need be presented for efficient synchronous modeling. In this dissertation, a multilevel representation based 3D synchronous modeling method is proposed.By analyzing design intent implied in 3D shape, the multilevel representation of CAD models is presented. Based on the hints of topology variation, face feature, generic shape feature and shape relation feature will be recognized to express simple design intent. According to the hierarchy of these features, a multilevel representation (MLR) of CAD model is introduced using a multi-tree structure. In the multi-tree, tree nodes correspond to features, and different combinations of tree nodes express diversified complex design intent.A method for extracting complex shape features that are difficultly defined by specific mode is presented based on heuristic shape matching. An enhanced attributed graph (EAG) is introduced to reflect the shape information of simple features in a MLR by enhanced attributes. The EAGs of feature template and target model are constructed by converting their MLRs. Then the corresponding sub-part of template is found out from target by graph matching. During graph matching, a tabu list of possible imp

您可能关注的文档

文档评论(0)

1亿VIP精品文档

相关文档