• ISSN: 1674-7461
  • CN: 11-5823/TU
  • 主管:中国科学技术协会
  • 主办:中国图学学会
  • 承办:中国建筑科学研究院有限公司

基于专用数据库的BIM数据管理AI智能体技术研究

BIM Data Management AI Agent Based on a BIM-specific Database

  • 摘要: 本文探讨了基于BIM专用数据库和AI大模型双引擎驱动的新一代BIM全生命期海量数据管理智能体技术。不同于基于模型文件的BIM数据管理模式,BIM专用数据库融合当前最先进的NewSQL及NoSQL技术,全面精确管理到每一个BIM构件、零件,包含属性/属性集、几何拓扑及关联关系等。BIM专用数据库对外提供统一的结构化查询SQL接口,使得全生命期应用的业务链调用变得简单。同时,基于AI大模型的文本理解能力,用户可基于自然语言与专用数据库直接进行交互,查询并获取所需BIM数据或子模型。基于BIM专用数据库及AI大模型加持的BIM数据管理智能体,可在协同设计、施工管理、碰撞检查、快速算量及资产管理等多种BIM应用场景中实现提质增效。

     

    Abstract: This article explores the next generation of BIM full-lifecycle massive data management agent technology, driven by a dual engine of BIM-specific database and AI big model. Different from the BIM data management mode based on model files, the BIM special database integrates the most advanced NewSQL and NoSQL technologies. The database is fully and accurately managed to each BIM component, part, including attribute / attribute set, geometric topology and association relationship. The BIM-specific database provides a unified structured query (SQL) interface to make business chain calls for full-lifecycle applications as simple as possible. Moreover, based on the text processing capabilities of the AI large model, users can directly interact with the dedicated database based on natural language to query and obtain the required BIM data or sub-models. The BIM data management agent, powered by the BIM-specific database and AI large model, can improve quality and efficiency in a variety of BIM application scenarios, including collaborative design, construction management, collision detection, fast quantity calculation, and asset management.

     

/

返回文章
返回