Research On Architectural Space Topology Extract And Retrieval Based On BIM

The emergence of BIM launched a digital revolution of information technology in the construction industry. Based on a three-dimensional digital technology, BIM took advantage of the three-dimensional computer modeling tools creating an integrated construction projects including all the digital model information. Thus,BIM could provide coordinated and internal unified digital information for design, construction, operation and maintenance. For the digital model with abundant data, its efficient reuse has important application value and practical significance.For the architects, the design of programs to retrieve reuse is an effective means of improve building design efficiency. Based on the features of contemporary architectural digitized design and with the help of computer technology, they can simplify the architectural design process effectively by retrieving the existing models and mining material for inspiration or reuse. However, the widely used retrieval method based on keyword search relies on the describing text only. Besides, it cannot take advantage of the complex and diverse forms of architectural space relationship and other information effectively. Therefore, it’s particularly necessary to construct a content-based retrieval structure and use BIM model as a search condition to get a similar architectural design case. In addition, BIM model is a three-dimensional building model, but the conventional three-dimensional shape retrieval methods rely mainly on global or local geometric features and shape description, so it is not fully applicable for the architectural models. For the above methods cannot effectively use the building structure of space, this paper proposes a retrieval method based architectural space topology. The main work is as follows:1. This paper deeply analyzed BIM, the architectural space combination forms, space syntax theory and graph matching method, studied and mastered the typical categories architectural spatial relationship(ARS) and their characteristics. Based on space syntax theory and graph theory, we proposed the RCTG model to formally describe the ARS and store the building information modeling’s space syntax variable attribute information, so as to provide data base for the retrieval method’s implement.2. The thesis analyzed the BIM model’s component information related to the communication and isolation between the various spaces. By extracting the core component, the BIM model internal space’s communicating relationship can be expressed with topological maps and the topological relation extraction algorithm can be constructed as well. We studied Revit software’s modeling method, had a good command of Revit software’s secondary development processes and achieved the building information model RCTG’s automatic extraction plug using C# language on the Visual Studio 2010 platform.3. Combining space syntax variables in the space syntax theory, the RCTG’s room elements were coded by their IDs and the model morphological analysis variables’ node depth value table(NDVT) was proposed as well. Thus it laid the database for the part of filtering and obtaining candidate model in the retrieval procedure.4. Based on space syntax theory, BIM model was abstract into mathematical model, thus an model feature vector embedded with space syntax variable properties was defined to provide data base for the model retrieval.5. This paper designed and developed a system for building information model retrieval according to the proposed retrieval method based on MFC application framework in Visual C ++ 6.0 platform. This system could be used in conjunction with RCTG extraction plug to implement the content-based retrieval using BIM model.The proposed method can retrieve the BIM models effectively and provide basic method and technique support for the construction of intelligent BIM-based CAD platform.