多高层模块化建筑布局智能生成技术-《建筑科学与工程学报》

文章信息/Info

Title:: Intelligent layout generation technology for multistory modular building

作者:: 刘界鹏，王俐强，夏毅，齐宏拓，陈杰，李江，李嘉琛; 1. 重庆大学土木工程学院，重庆 400045； 2. 重庆大学山区土木工程安全与韧性全国重点实验室，重庆 400045； 3. 香港科技大学香港智能建造研发中心，香港 999077；4. 中建科工集团绿色科技有限公司，广东深圳 518107

Author(s):: LIU Jiepeng, WANG Liqiang, XIA Yi, QI Hongtuo, CHEN Jie, LI jiang, LI Jiachen; 1. School of Civil Engineering, Chongqing University, Chongqing 400045, China; 2. State Key Laboratory of Safety and Resilience of Civil Engineering in Mountain Area, Chongqing University, Chongqing 400045, China; 3. Hong Kong Center for Construction Robotics, The Hong Kong University of Science and Technology, Hong Kong 999077, China; 4. China Construction Science and Industry Group Green Technology Co., Ltd., Shenzhen 518107, Guangdong, China

关键词:: 模块化建筑; 深度强化学习; 模块布局; 生成式设计; 平台开发

Keywords:: modular building; deep reinforcement learning; modular layout; generative design; platform development

分类号:: -

DOI:: -

文献标志码:: A

摘要:: 针对模块化建筑布局设计中人工设计困难和自动化程度低等问题，提出了多高层模块化建筑智能布局设计方法，实现了多样性布局方案的高效与自动生成。首先针对模块化建筑图纸进行自动信息提取与建筑区域分组，完成建筑初始二维图纸信息到标准信息格式的转换。基于深度强化学习对多高层模块化建筑进行布局方案智能生成，通过模块单元总数、房间调整值和房间调整量等奖励值的协同作用，实现模块单元布局方案的优化与评估；最后，为了提升工程实用性和操作便捷性，集成建筑信息提取、建筑布局设计、布局方案可视化等功能模块，搭建了模块化建筑智能布局设计平台。以立新湖学校项目为案例，采用所提出方法的3种预设规格的模块单元进行了布局方案的生成，评估了生成方案的模块单元总数、房间调整量、房间调整值3个指标以及归一化后的总指标值。结果表明：所提出的方法可以高效生成适用于不同场景的模块方案，提升了模块化建筑设计的自动化与智能化水平，为工程实践提供了技术支撑。

Abstract:: Aiming at the difficulties of manual design and low degree of automation in modular building layout design, an intelligent layout design method for multistory modular buildings was proposed, which achieved efficient and automatic generation of diverse layout schemes. Firstly, automatic information extraction and building area grouping were carried out for modular architectural drawings, completing the conversion of initial twodimensional drawing information to standard information format. The layout schemes for multistory modular buildings were intelligently generated based on deep reinforcement learning. Through the synergistic effect of reward values such as the total number of module units, room adjustment values, and room adjustment quantities, the optimization and evaluation of module unit layout schemes were achieved. Finally, in order to enhance the practicality and operational convenience of the project, a modular intelligent building layout design platform was built by integrating functional modules such as building information extraction, building layout design, and layout scheme visualization. Taking the Lixinhu School project as study case, layout schemes were generated by using three preset specification module units in the proposed method. The total number of module units, room adjustment amount, room adjustment value, and normalized total indicator value for generating the plan were evaluated. The results show that the proposed method can efficiently generate module schemes suitable for different scenarios, improve the automation and intelligence level of modular building design, and provide technical support for engineering practice.

《建筑科学与工程学报》[ISSN:1673-2049/CN:61-1442/TU]

文章信息/Info

参考文献/References:

相似文献/References:

备注/Memo

常用功能

导航/Navigate

工具/Tools

统计/Statistics