为辨别不同时段不同地区影响单车出行的关键因素，认识共享单车与用户间的动态平衡关系，基于对城市多源数据的分析，从用车环境、道路交通基础设施、公共交通变量和土地利用变量4方面选取共享单车出行影响因素，并以此为输入性指标建立全局回归分析初选模型与地理加权回归模型（Geographically Weighted Regression, GWR），初步分析各因子的空间非平稳性。通过地理差异检验得到具有空间异质性的影响因子，然后采用空间可视化分析方法进一步探究单车出行分布与关键因子间的作用机理。以西安市建成区共享单车出行数据为例进行分析，结果表明：在工作日早高峰和晚高峰时段，共享单车设施分布密度对共享单车出行起点分布的影响存在空间异质性，企业密度在工作日早高峰时段、餐饮类分布密度在工作日晚高峰时段对共享单车出行终点的影响具有空间异质性；GWR模型对单车起终点影响因素回归分析的拟合优度R2相比全局回归模型分别提高了0.428, 0.378, 0.208 和0.309，且修正后的赤池信息准则（Akaike Information Criterion, AIC）值（AICc）与残差平方和均有所降低，表明GWR 模型能较好地反映变量之间的空间异质性，具有较高的准确性与适用性。
In order to identify the key factors affecting bicycle travel in different regions at different time and understand the dynamic balance relationship between bike sharing and users, based on the analysis of urban multi-source data, the influencing factors of bike sharing travel were chosen from four aspects: bike sharing environment, road traffic infrastructure, public transportation and land use variables. A Global Regression model and a Geographic Weighted Regression(GWR) model were established taking those influencing factors as input indicators. The spatial non-stationarity of each factor was analyzed preliminarily. Through the geographical difference test, the influencing factors with spatial heterogeneity were obtained. Then the interaction mechanism between the bicycle trip distribution and the key factors was further explored via using the spatial visualization analysis method. The bike sharing travel data in the built-up areas of Xi'an was analyzed. The results showed that in the morning and evening peaks of working days, the impact of the distribution density of bike sharing facilities on the distribution of the starting points of bicycle trips had a spatial heterogeneity, the impact of the density of enterprises on the end points of bicycle trips during morning peaks of working days and the impact of the density of catering services on the end points of bicycle trips during evening peaks of working days had a spatial heterogeneity; compared with the Global Regression model, the goodness of fit of GWR Model for regression analysis of influencing factors of the start and end points of bike sharing increased by 0.428, 0.378, 0.208 and 0.309 respectively, and the value of AICc and the sum of squared residuals reduced. It shows that GWR model can describe the spatial heterogeneity of variables well, and has high accuracy and applicability.
[1] Vogel M, Hamon R, Lozenguez G, et al. From Bicycle Sharing System Movements to Users: a Typology of Vélo′v Cyclists in Lyon Based on Large-Scale Behavioural Dataset[J]. Journal of Transport Geography, 2014, 41: 280-291.
[2] 莫海彤，魏宗财，翟青. 老城区共享单车出行特征及影响因素研究——以广州为例[J]. 南方建筑，2019（1）：7-12.
[3] 郭海富，喻永生，罗瑜. 基于共享单车时间分布优化调度模型[J]. 科技视界，2018（7）：101.
[4] 周荣，王元庆，朱亮，等. 基于时空数据的共享单车出行特征研究[J]. 武汉理工大学学报（交通科学与工程版），2019，43（1）：159-163.
[5] 魏宗财，莫海彤，刘玉亭. 面向智慧城市的共享单车出行时空间特征研究——以广州天河中心区为例[J]. 科技导报，2018，36（18）：71-80.
[6] 陈奇放，付泽钰. 共享单车影响下居民骑行行为时空特征研究——以南京市中心城区为例[C]// 2018中国城市规划年会论文集（05城市规划新技术应用）. 杭州：中国城市规划学会，2018：853-861.
[7] 杨蒙，陈天，臧鑫宇，等. 基于智慧数据的共享单车聚集特征与街道改造策略研究——以天津市和平区为例[J]. 现代城市研究，2019（6）：9-15.
[8] Contardo C, Morency C, Rousseau L M. Balancing a Dynamic Public Bike-Sharing System[R]. Montreal, Canada: CIRRELT, 2012.
[9] 罗桑扎西，甄峰，尹秋怡. 城市公共自行车使用与建成环境的关系研究——以南京市桥北片区为例[J]. 地理科学，2018，38（3）：332-341.
[10] Faghih-Imani A, Eluru N, EI-Geneidy A M, et al. How Land-Use and Urban form Impact Bicycle Flows: Evidence from the Bicycle-Sharing System (BIXI) in Montreal[J]. Journal of Transport Geography, 2014, 41: 306-314.
[11] Tran T D, Ovtracht N, D′Arcier B F. Modeling Bike Sharing System Using Built Environment Factors[J]. Procedia CIRP, 2015, 30: 293-298.
[12] Faghih-Imani A, Eluru N. Incorporating the Impact of Spatio-Temporal Interactions on Bicycle Sharing System Demand: A Case Study of New York CitiBike System[J]. Journal of Transport Geography, 2016, 54: 218-227.
[13] Cervero R, Sarmiento O L, Jacoby E, et al. Influences of Built Environments on Walking and Cycling: Lessons from Bogotá[J]. International Journal of Sustainable Transportation, 2009, 3(4): 203-226.
[14] EI-Assi W, Mahmoud M S, Habib K N. Effects of Built Environment and Weather on Bike Sharing Demand: a Station Level Analysis of Commercial Bike Sharing in Toronto[J]. Transportation, 2017, 44(3): 589-613.
[15] Handy S L, Boarnet M G, Ewing R, et al. How the Built Environment Affects Physical Activity: Views from Urban Planning[J]. American Journal of Preventive Medicine, 2002, 23(2): 64-73.
[16] Corcoran J, Li T, Rohde D, et al. Spatio-Temporal Patterns of a Public Bicycle Sharing Program: the Effect of Weather and Calendar Events[J]. Journal of Transport Geography, 2014, 41: 292-305.
[17] 曾小利，陈宇，钱志诚. 基于摩拜数据的居民单车通勤特征分析——以深圳市南山区为例[C]// 2018中国城市规划年会论文集（06城市交通规划）. 杭州：中国城市规划学会，2018：903-912.
[18] 李煜茜，张能，王珊珊，等. 单车上的老北京——北京老城区骑行行为影响因素研究[J]. 北京规划建设，2018（3）：6-12.
[19] 高枫，李少英，吴志峰，等. 广州市主城区共享单车骑行目的地时空特征与影响因素[J]. 地理研究，2019，38（12）：2859-2872.
[20] Brunsdon C, Fotheringham A S, Charlton M E. Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity[J]. Geographical Analysis, 1996, 28(4): 281-298.
[21] 叶显严. 基于地理加权回归的城市出租车客流影响因素分析及建模[D]. 成都：西南交通大学，2017.
[22] 张俊杰. 基于地理加权回归模型公交客流影响因素的空间异质性研究[D]. 西安：长安大学，2018.
[23] Zhang Y, Thomas T, Brussel M, et al. Exploring the Impact of Built Environment Factors on the Use of Public Bikes at Bike Stations: Case Study in Zhongshan, China[J]. Journal of Transport Geography, 2017, 58: 59-70.
[24] Cui Y, Mishra S, Welch T F. Land Use Effects on Bicycle Ridership: A Framework for State Planning Agencies-Science Direct[J]. Journal of Transport Geography, 2014, 41: 220-228.
[25] Kobayashi T, Lane B W. Spatial Heterogeneity and Transit Use[C]// Proceedings of 11th World Conference on Transport Research. Berkeley CA: 11th World Conference on Transport Research, 2007.
[26] Sung H, Oh J T. Transit-Oriented Development in a High-Density City: Identifying its Association with Transit Ridership in Seoul, Korea[J]. Cities, 2011, 28(1): 70-82.
[27] 覃文忠. 地理加权回归基本理论与应用研究[D].上海：同济大学，2007.
[28] Ma X, Cao R, Jin Y. Spatiotemporal Clustering Analysis of Bicycle Sharing System with Data Mining Approach[J]. Information, 2019, 10(5): 163.
[29] 张萌. 政策驱动下的土地利用空间布局变化模拟分析——基于CLUE-S土地利用变化模拟模型[D]. 昆明：昆明理工大学，2013.
京公网安备 11010102003417号
京ICP备17006742号-3