|
| 应急情景下基于强化学习的大空间建筑室内布局生成设计研究——以床位布局设计为例 |
| Research on the Generative Design of Interior Layout of Large-space Buildings Based on Reinforcement Learning in Emergency Scenarios- Take Bed Layout Design as an Example |
| 投稿时间：2025-05-07 修订日期：2025-12-05 |
| DOI： |
| 中文关键词: 大空间建筑、平急转换、深度强化学习、建筑生成设计、床位布局设计 |
| 英文关键词: Large-space buildings, Normal-to-emergency conversion, Deep reinforcement learning, Generative design of buildings, Bed layout design |
| 基金项目: |
|
| 摘要点击次数: 14 |
| 全文下载次数: 0 |
| 中文摘要: |
| 城市需要有高安全度、强应变能力的应急手段抵御自然灾害侵袭,研究和实践表明大空间建筑有助提升城市防灾韧性。针对应急情景的室内布局设计是利用大空间建筑进行防灾的关键任务,床位布局设计是其核心内容。本研究以床位布局设计为例讨论实现大空间建筑室内布局生成设计的方法和路径。首先,明确强化学习对布局设计问题的适用性。其次,按照计算机语境转译布局设计问题。然后,论述布局设计内容与强化学习算法的深度拟合方式。在此基础上,以床位布局生成设计模型的建立阐述强化学习算法的应用,并讨论人工智能算法融入布局设计的关键要素。本研究能为应急情景下建筑空间布局的生成设计提供方法参考,也能为建筑设计中运用强化学习技术提供思路。 |
| 英文摘要: |
| Cities need emergency measures with high safety and strong adaptability to resist natural disasters. Researches and practices have revealed that large-space buildings are conducive to improving the resilience of cities to disasters. The design of interior layout for emergency scenarios is a key task for disaster prevention using large-space buildings, and the bed layout design is at its heart. This study, with the bed layout design as an example, discusses the methods and paths to realize the generative design of the interior layout of large-space buildings. Firstly, the applicability of reinforcement learning to layout design problems is clarified. Secondly, the layout design problems are translated in the context of computers. Then, the deep fitting method between the layout design content and the reinforcement learning algorithm is discussed. On this basis, the establishment of a design model for bed layout generation is elaborated to illustrate the application of reinforcement learning algorithms, and the key elements of integrating artificial intelligence algorithms into layout design are discussed. This study can not only provide a method reference for the generative design of architectural spatial layout in emergency scenarios, but also put forward ideas for the application of reinforcement learning technology in architectural design. |
|
View Fulltext
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
