人本视角下的社区建成环境与老年人活力出行研究OACHSSCD

With the intensifying aging of the global population,improving the health of the elderly has become a core issue in urban development.As a sustainable form of physical activity that can be easily integrated into daily life,active travel(e.g.walking and cycling)is a critical pathway to realise the goal of"active aging".However,existing studies mainly rely on macro-scale GIS(Geographic Information System)data,which limits their ability to capture human-scale micro-environmental perceptions.In particular,insufficient attention has been paid to key indicators representing spatial openness and walkability,such as the sky view index and sidewalk index.Based on data from the Chengdu Comprehensive Travel Survey in 2016,the total daily active travel time of the elderly was chosen as the dependent variable,while gender,retirement status,household size and household income were chosen as the control variables.Community-level built environment indicators were measured by integrating multi-source big data,including open-source demographic data,road networks,points of interest(POIs)data,and street-view imagery(SVI).Specifically,SVI data were collected from Baidu Maps(2015～2017).Vegetation,sky,and sidewalk elements were recognised by using the HRNet semantic segmentation model,and three core indicators(the green view index,sky view index,and sidewalk index)were calculated.The multiple linear regression model was applied,and multicollinearity risks were eliminated in advance through the Pearson correlation test.The age of the elderly was adjusted from 60+to 65+for the robustness test.A total of 10,747 valid samples were included(including 6,631 robustness samples). The regression results indicate that the control variables of male,retired status,and high household income have significantly positive correlations with the active travel time of the elderly,while household size shows a significantly negative correlation.With respect to the built-environment variables,population density,the green view index,the sidewalk index,and accessibility to medical facilities have significantly positive effects on active travel.In contrast,road density and accessibility to bus stops have significantly negative effects.These reflect the travel risks of complex road networks and the"substitution effect"of public transport.Additionally,land-use mix and sky view index are not statistically significant.The robustness test confirms that both the influencing directions and significance levels of the core built-environment indicators remain stable. This study confirms that using human-centred street view indicators as an effective complement to the traditional community-level built environmental index system has important values for analysing the active travel behaviours of the elderly.Based on the regression analysis results,a strategy-generation framework of"empirical evidence-underlying mechanisms-planning responses"was built,and the strategies of"promote beneficial factors and curbing adverse factors"were proposed according to the dual criteria of statistical significance and spatial intervenability.Firstly,it shall create a high-quality street environment in communities by expanding sidewalks,adding cycling lanes,and optimising the green view index.Secondly,it shall optimise travel routes for medical access and improve the multi-level community medical system by building a"15-minute medical service circle".Thirdly,it shall build high-efficiency community-level road networks by implementing traffic calming measures,perfecting road signs,and installing safety islands.The"street-medical-road network"coordinated optimisation mode not only provides empirical guidance for age-friendly community renewal in Chengdu but also offers a reference for global high-density cities to cope with aging challenges.