摘要
城市轨道中心型影响区的土地利用规划是TOD研究中的重要内容。本文根据站点影响区"节点—场所"的双重特性确定土地利用指标因子,设计了基于GIS的站点影响区土地利用数据库框架,进而提出了"在较少样本条件基于BP人工神经网络的站点影响区土地利用预测模拟"和"在大量样本条件下基于深度神经网络的站点影响区土地利用规划方案生成"两条技术路线,探索了运用人工智能技术进行站点影响区土地利用研究的跨学科方法,为站点影响区土地高效利用和集约发展提供了可行的技术方法和研究思路。
According to the dual characteristics of"node-place"in the affected area of urban rail central station,this paper determined the land use index factor,and designed the land use database framework of the station affected area based on GIS.Furthermore,two technical routes were proposed:"simulation of land use prediction in site-affected area based on BP artificial neural network under fewer sample conditions"and"generation of land use planning plan for station affected area based on deep neural network under a large number of samples".This research explores the interdisciplinary method of using artificial intelligence technology to study the land use of the station affected area,and provides feasible technical methods and research ideas for the efficient use of land and intensive development in this area.
作者
何媛
袁红
宋秋明
武子栋
He Yuan;Yuan Hong;Song Qiuming;Wu Zidong(School of Architecture and Design,Southwest Jiaotong University,Chengdu Sichuan 611756,China;Chongqing Institute of Landscape Architecture and Planning,Chongqing 400000,China;General Institute of Architectural Planning and Design,Chongqing University,Chongqing 400045,China)
出处
《城市建筑》
2021年第1期39-41,57,共4页
Urbanism and Architecture
基金
国家自然科学基金面上项目“山城轨道影响区地下空间立体紧凑设计理论研究-基于大数据及3D GIS技术”(51678486)。
