课题基金 / 基金详情

融合轻量化网格与B-Rep的云架构CAD软件特征模型数据表示方法研究

批准号:
52105291
项目类别:
青年科学基金项目(C类)
资助金额:
30.0 万元
负责人:
赵飞宇
依托单位:
学科分类:
机械设计学
结题年份:
2024
批准年份:
2021
项目状态:
已结题
项目参与者:
赵飞宇

项目摘要

结项摘要

相似基金

相关文献

中文摘要
云架构CAD软件对于推动工业技术云端软件化至关重要。当前云架构CAD软件受制于Web环境下数据传输与渲染效率,普遍采用数据表示局限于三角面片顶点与边结构几何与拓扑关系的轻量化网格模型,难以满足机械产品多特征语义耦合、约束驱动的参数化特征建模需求。本项目探索使轻量化网格与B-Rep融合且B-Rep约束下网格局部变形的云架构CAD软件特征模型数据表示方法。重点研究基于离散流形网格多微分特性耦合分析的边界形面辨识;以重构B-Rep拓扑关系驱动的保特征网格重划分,基于邻接形面干涉分析建立模型B-Rep拓扑关系并驱动特征语义识别与网格重划分;基于特征语义驱动且维持B-Rep拓扑关系的网格局部变形原理,建立边界约束的特征语义构形矢量参数计算模型,并提出参数驱动的网格局部变形算法。项目旨在突破基于Web的三维数字化建模基础核心技术瓶颈,对于我国研发自主可控的云架构CAD软件具有重要学术意义和应用价值。
英文摘要
Cloud CAD software is essential to promote the cloud softwareization of industrial technology. However, it is restricted by the efficiency of data transmission and rendering in the Web environment that most of the current cloud CAD softwares still use the lightweight triangular mesh, and the topological connection is limited to the relationship between the vertices and edges of the triangular faces, which are difficult to satisfy the requirements of the parametric 3D feature modeling of the mechanical products driven by the combination and constraint of the multi-features. This project explores the data representation method of feature model for cloud CAD software that integrating lightweight mesh with B-Rep and B-Rep constrained mesh local deformation. The boundary shape identification based on coupling analysis of multi-differential characteristics of discrete manifold mesh is studied. The feature-preserving remeshing method driven by reconstructing B-Rep topological relations is also studied, the topological relations of 3D mesh are built based on the interference analysis of adjacent surfaces, which promote the feature recognition and remeshing. The feature driven and B-Rep constrained mesh local deformation principle is also studied, a boundary constrained shape parameter calculation mode of feature is built, and a parameter driven local mesh deformation algorithm is proposed. Therefore, the research contributes to break through the core technology bottleneck of Web-based digital 3D modeling, which is of great academic and application significance for the research and development of independent and controllable cloud CAD software in China.
云架构计算机辅助设计(CAD)软件在推动工业技术云端化方面具有重要意义。然而,当前云架构CAD软件的数据表示普遍局限于三角面片顶点和边结构的轻量化网格模型,这种表示方式难以满足机械产品多特征语义耦合与约束驱动的参数化特征建模需求。本项目旨在探索轻量化离散流形网格与边界表示法(B-Rep)相融合的新型CAD模型数据表示方法。项目首先对云架构CAD软件的关键技术与应用进行了全面综述,深入分析了现有技术的局限性,并提出了改进思路。接着,研究了基于焦点损失和正则化损失的MeshNet三维形状分类性能优化方法,从深层次挖掘轻量化网格模型底层蕴含的特征语义;进一步,探索了基于分层神经网络的点云模型分割与几何基元拟合方法,为CAD模型的自动特征重构提供了新途径;在此基础上,本项目提出了一种用于从分割的流形三角网格模型中重建实体CAD特征的新型计算范式,并在Web环境下研究了多参数驱动的复杂几何体三维轻量化建模方法,具有理论指导意义。项目还通过数字孪生车间设备的轻量化网格模型装配特征构建进行了应用验证,展示了该方法在云原生工业软件和数字孪生建模等领域的潜在应用前景。.通过本项目的研究,已在《Computers & Graphics》、《计算机集成制造系统》等国际与国内知名期刊上发表论文6篇,其中SCI/EI收录论文5篇;授权发明专利2项;授权软件著作权1项;开发了云架构CAD软件原型系统1套;培养青年教师1名,培养在读硕士研究生5名;作为第一指导教师指导的学生获得科技竞赛全国总决赛奖项2项。
国内基金
海外基金