基于生成对抗网络(GAN)和NSGA-2遗传算法的汉口滨江居住区采光优化研究  

Study on the lighting optimization of Hankou riverside residential block based on Generative Adversarial Network(GAN)and NSGA-2 genetic algorithm

作　　者：王孝鑫 李竞一 Wang Xiaoxin;Li Jingyi

机构地区：[1]武汉城市学院城建学部 [2]华中科技大学 [3]多维建筑设计工作室

出　　处：《建筑技艺》2021年第9期84-88,共5页Architecture Technique

摘　　要：随着人工智能技术在各个领域的广泛运用,越来越多的设计人员开始尝试将人工智能技术的成果运用到城市或建筑设计当中。通过汉口滨江居住区城市数据和人工智能技术控制区域三维模型的合理生成,并对整体区域建筑环境进行环境模拟,达到居住区布局及造型的优化设计的目的,最后通过优化设计案例为设计师提供设计建议。With the application of artificial intelligence technology in various fields,more and more designers try to apply it to urban or architectural design.Through the urban data and artificial intelligence technology of Hankou riverside residential area,the reasonable generation of regional three-dimensional model can be controlled,and the environmental simulation of the overall regional architectural environment can optimize the design of residential area layout and modeling.Finally,design suggestions can be provided through optimization design cases.

关 键 词：深度学习 生成对抗网络 NSGA-2遗传算法 居住区改造 日照模拟 优化设计 

分 类 号：TU201.4[建筑科学—建筑设计及理论]