南京民国建筑修缮BIM模型实例库的构建及其数据挖掘与知识发现研究

The buildings constructed in the period of the Republic of China in Nanjing are valuable heritages of Chinese modern architecture. The rehabilitation of the demolished buildings is an important means of heritage protection. But the archived documents of the most repaired projects are still two-dimensional drawings，lower data integration of two-dimensional drawings can not fully meet the needs of the architectural heritage conservation in the entire life cycle. The key Laboratory of Urban and Architectural Heritage Conservation of Ministry of Education (Southeast University) has completed many rehabilitation designs of MinGuo buildings in Nanjing over the years. By selecting the appropriate sample from the existing data，the study will explore the application of BIM models in the rehabilitation design of architectural heritages and the construction procedure of BIM models with professional modification of the Autodesk Revit BIM Software's tectonic template used in MinGuo architectural heritages.Relied on metadata indexing，the research is aiming at the construction of BIM models' library coupling with the domain expert's design experience and conservation knowledge about sample architectural heritages. Further , with the aid of the tool for data mining within BIM models, the relation between the evolution routine of MinGuo building techniques and the variance of modern rehabilitation techniques will be dug out .

南京民国建筑是中国近代建筑的宝贵遗产，对其进行维护修缮更新是遗产保护的重要手段，但目前大多数项目修缮完成后存档的仍旧是二维图纸，数据集成度低，不能充分满足建筑遗产完整生命周期的信息供给需要。课题以东南大学建筑学院暨东南大学城市和建筑遗产保护教育部重点实验室历年承担的南京民国建筑保护修缮设计项目为主要资源，从中选择适量样本, 探索BIM模型在近代建筑遗产保护中的应用，通过对BIM软件revit的族库二次开发和针对建筑遗产专业需求的属性扩展，构建了相关建筑的BIM模型，课题基于所建BIM模型，通过设计建筑元数据标引，耦合项目修缮设计者作为领域专家的案例经验知识，构成具有知识图底的BIM模型实例库，并尝试通过数据挖掘技术，将BIM模型库的数据关联到外部知识库进行知识发现，探索从精细的BIM数据层面挖掘南京近代建筑技术演进中的各种变异性态及其与现代修缮技术之间的适配性关联。

本课题对南京民国建筑典型项目进行BIM模型创建，探索提高相关文档管理的数据集成度和可视度以满足建筑遗产完整生命周期的信息供给需求；课题利用BIM REVIT软件的专用族创建机制，完成了民国建筑修缮改造的相关族库，同时基于最新发展的WEB gl技术，将相关HBIM模型转化到web环境，基于建筑遗产元数据整合相关的多类型数据源以及外部数据库，提供知识挖掘入口以形成多维度的知识矩阵， 探索从BIM数据层面创建知识关联，为历史建筑的信息存放提取提供了更丰富的路径，在建筑遗产保护的信息管理领域有一定的应用示范价值。