摘要
结合形体设计的特点提出了一种建筑形体性能评价与修改辅助方法构想,并给出了结合参数化模型、性能模拟、机器学习的技术路线;而后以音乐厅及其音质性能为例开发了工具雏形、实现了构想的方法;最后对工具雏形进行了应用测试。结果表明,给出的方法和工具雏形,具有较好的可靠性和反馈实时性;适用于形体设计,能够帮助建筑师在不了解技术知识且没有顾问辅助下,将形体从性能不理想逐步改进至理想,有助于提高形体设计的性能及建筑师的工作效率。
This paper advocates for the performance evaluation and modification suggestion method of the building form based on characteristics of formal design, and offers a technology roadmap that integrates parametric model, performance simulation and machine learning.It develops a rudimentary tool and ways of realization for concert halls and acoustic performance within them.Tests of the tool verifies that the method and tool are reliable with real-time feedback, and they are applicable in formal design to help architects gradually optimize the form for better building performance, with either no sufficient technical knowledge or assistance of technical consultants. hence the improvement of building performance and working efficiency of architects.
作者
吕帅
李靖宇
李珺杰
LYU Shuai;LI Jingyu;LI Junjie
出处
《建筑学报》
CSSCI
北大核心
2019年第2期31-37,共7页
Architectural Journal
基金
国家自然科学基金青年基金项目(51708355)
亚热带建筑科学国家重点实验室开放课题项目(2019ZB14)
北京市自然科学基金项目(8182043)
深圳大学新引进教师科研启动项目(2018072)
关键词
数字设计
形体设计
机器学习
性能潜力
修改辅助
digital design
formal design
machine learning
performance potential
modification suggestion
