摘要
探析城市街道空间品质与居民出行行为的关联机制对合理有效配置街道设施资源、促进健康绿色出行具有重要意义。以合肥市七桂塘街区为例,通过Python爬取街景图像,采用机器学习算法、空间句法结合ArcGIS量化街道空间品质构成要素;通过行为观察获取出行数据,进而构建空间品质和行为特征关联研究多元线性回归模型总结各影响因子相互作用程度和方式。研究表明:出行行为与街道空间要素存在特定的数理影响关系,其中功能业态、步行宽度和界面开放度对出行行为影响更显著。在此基础上提出老城区街道优化策略,以期为合肥市街道设计导则的制定提供借鉴。
The new data environment lays a foundation for quantitative analysis of correlation between street space and behaviours.With the continuous urbanisation of China,the vehicle-oriented urban spatial expansion mode leads to a lack of pedestrian space for residents.This issue triggered rethinking and improvement of research into street spaces.The topics of improvement and design of street spaces are attracting increased academic attention.As a new type of special research into road spaces,street design guidelines are increasingly prominent in China and foreign cities.Exploring the correlation mechanism between urban street spaces and the travel behaviours of residents has important significance for the reasonable and effective allocation of street facility resources,for promoting health activities and green travel,and for enriching the depth of street guidelines.Qiguitang Street in the old urban area of Hefei was chosen to collect empirical data.Firstly,OSM was used to crawl street images through Python.Components of street space quality were quantified by combining machine learning algorithms,space syntax,and the ArcGIS platform.The construction status of urban street spaces was analysed.Secondly,data about travel behaviours was collected through behaviour observation and the spatial-temporal characteristics of behaviours in streets were analysed.Finally,a multivariate linear regression model to study correlations between four spatial qualities(comfort,safety,convenience,and richness)and pedestrian behavioural characteristics was constructed.From this,a corresponding index system with 12 variables was determined.The degree of interaction and interaction methods between influencing factors was summarised according to data processing results.The research demonstrated a mathematical influencing relationship between travel behaviours and street space elements.The specific behaviours of pedestrians may be influenced by specific environmental features,and environmental features may also guide pedestrians towards specific
作者
宣蔚
彭康
臧传云
XUAN Wei;PENG Kang;ZANG Chuanyun
出处
《南方建筑》
CSCD
北大核心
2023年第4期49-60,共12页
South Architecture
基金
国家自然科学基金资助项目(52008143):基于人迹地图数据的城市街区空间组织与居民低碳出行的关联机制研究
安徽省哲学社会科学规划项目(AHSKY2021D74)。
关键词
多源大数据
街道空间品质
行为特征
街景图片
multisource big data
street space quality
behavioural characteristics
streetscape images
