研究内容

Research contents

骑行作为一种健康且环保的交通方式，在促进人类健康和构建可持续城市方面发挥着重要作用。

骑行行为的普及受多种因素影响，其中建成环境是关键因素之一。

现有研究多依赖于问卷调查和访谈数据，缺乏与实时骑行数据的结合。

探讨建成环境特征与骑行行为之间的关联。

基于这些信息，提出改善城市骑行频率和体验的建议。

选择澳大利亚悉尼市中心作为研究区域，因其复杂的道路网络和建成环境。

Cycling, as a healthy and environmentally friendly mode of transportation, plays an important role in promoting human health and building sustainable cities.

The popularity of cycling is influenced by many factors, among which the built environment is one of the key factors.

Existing research mostly relies on questionnaires and interview data and lacks integration with real-time cycling data.

Exploring the association between built environment characteristics and cycling behavior.

Based on this information, recommendations are made to improve urban cycling frequency and experience.

Central Sydney, Australia, was chosen as the study area due to its complex road network and built environment.

研究方法

Research methods

街景图像数据：通过Google Street View API收集街景图像，并使用IPAB模型进行图像分割，提取23个建成环境特征。

骑行数据：使用Strava Metro Data（SMD）获取悉尼市2023年的骑行数据，包括骑行次数、时间和目的等。

对街景图像进行分割，提取绿化、水域视图、天空开阔度、封闭感等骑行感知指标。

对骑行数据进行处理，计算每月平均骑行次数，并与街景图像数据进行空间匹配。

使用多元泊松回归模型分析建成环境特征与骑行行为之间的关联。

分别分析通勤骑行和休闲骑行的差异。

Street View image data: Street View images were collected through the Google Street View API, and the IPAB model was used for image segmentation to extract 23 built 

environment features.

Cycling data: Use Strava Metro Data (SMD) to obtain Sydney City’s cycling data in 2023, including the number of rides, time and purpose.

Segment the street view image and extract cycling perception indicators such as greening, water view, sky openness, and sense of closure.

The cycling data is processed to calculate the average number of rides per month and perform spatial matching with Street View image data.

A multivariate Poisson regression model was used to analyze the association between built environment characteristics and cycling behavior.

Analyze the differences between commuting and leisure riding separately.

研究结果

Research conclusions

正面关联：街道绿化、自行车道、交通信号灯和路边停车等因素与骑行频率呈正相关。

负面关联：天空开阔度、封闭感和交通标志框等与骑行频率呈负相关。

高骑行频率道路主要集中在主干道，而次干道和支路骑行频率较低。

通勤骑行主要集中在市中心和公交站点附近，休闲骑行则更多围绕城市绿地和水域。

市中心绿化和水域视图较低，而郊区较高。

市中心天空开阔度较低且封闭感较强，而郊区相反。

绿化对通勤骑行和休闲骑行均有正面影响。

水域视图对休闲骑行有正面影响，但对通勤骑行影响较小。

天空开阔度和封闭感对两种骑行目的均有负面影响。

Positive correlations: Factors such as street greening, bike lanes, traffic lights and on-street parking are positively correlated with cycling frequency.

Negative correlations: Sky openness, sense of closure, and traffic sign frames are negatively correlated with cycling frequency.

Roads with high cycling frequency are mainly concentrated on main roads, while secondary roads and branch roads have low cycling frequency.

Commuting cycling is mainly concentrated in the city center and near bus stations, while leisure cycling is more centered around urban green spaces and water bodies.

Downtown greenery and water views are lower, while suburban areas are higher.

The sky is less open and closed in the city center, while the opposite is true in the suburbs.

Greening has a positive impact on both commuting and recreational cycling.

Water views have a positive impact on recreational riding but have a smaller impact on commuter riding.

Sky openness and a sense of confinement have a negative impact on both riding purposes.