| 663 | 1 | 316 |
| 下载次数 | 被引频次 | 阅读次数 |
【背景】地铁作为中大型城市中长距离出行的重要公共交通工具可以有效解决城市居民的主要通勤需求问题，共享单车作为一种新兴的短距离出行方式，有效填补了“最后一公里”的出行空白，两者之间的相互作用模式直接关系到居民的出行选择以及城市交通的整体运行效率。【目标】探讨如何通过优化建成环境促进共享单车-地铁之间更加和谐且有效的合作与竞争关系(延伸出行、竞争出行、补充出行)，从而提升城市交通系统的整体效能和可持续性。【方法】从土地利用特征、交通附属设施属性、社会经济属性等方面，利用停车场数量、土地利用香农熵、平均房价、人口数量等17个建成环境指标作为自变量，以三种出行模式下的共享单车出行量为因变量，构建了基于随机搜索优化超参数的XGBOOST模型，并用SHAP可解释模型进行变量分析。【结果】模型评估结果显示，延伸出行和竞争出行的模型测试集R2分别为0.84和0.82，补充出行的模型R2为0.76。【结论】SHAP进一步揭示了停车场数量、到市中心距离和人口数量对共享单车-地铁竞合关系的非线性影响。停车场数量的影响总体上呈现先降低后上升的趋势，延伸出行模式较为均衡，补充出行模式呈S型增长，达到一定数量后边际效应减弱。到市中心的距离对共享单车使用的影响表现为先增加后减少，延伸出行模式在20 km内需求较高，竞争出行模式随距离增加而衰减，补充出行模式在15 km外接驳需求下降。人口数量对延伸出行和补充出行模式呈S型增长，高密度区域共享单车需求显著增加，而竞争出行模式受人口影响较小，更依赖地铁等公共交通方式。【应用】研究结果为城市交通规划提供参考，特别在停车场较少的区域，可以通过增加共享单车投放来提升短途接驳效率；在20 km以内区域，加强共享单车与地铁接驳服务，有助于优化换乘体验。合理调整建成环境，可提升共享单车使用率，减少对私家车的依赖，降低交通碳排放，促进绿色出行和可持续发展。
Abstract：[Background] As an important mode of public transportation in medium-to large-sized cities, subways effectively address the primary commuting requirements of urban residents over medium to long distances. Meanwhile, bike-sharing, as an emerging short-distance travel mode, provides an essential solution for bridging the “last mile” gap. The interaction patterns between these two modes directly affect the travel choices of residents and the overall efficiency of urban transport systems. [Objective] This study aims to explore how optimizing the built environment can promote a more harmonious and effective cooperative-competitive relationship between bike-sharing systems and subways, which can be categorized into three typical patterns: extension travel, competitive travel, and supplementary travel, thereby enhancing the overall efficiency and sustainability of the urban transportation system. [Method] Based on land-use characteristics, transportation facility attributes,and socioeconomic attributes, 17 built environment indicators, including the number of parking lots,Shannon entropy of land use, average housing price, and population size, are used as independent variables, whereas shared-bike ridership under the three travel modes is adopted as the dependent variable. An XGBOOST model optimized through a random search for hyperparameter tuning was established, and SHAP interpretability analysis was employed for variable importance analysis.[Result] The model evaluation results indicate that the R2 values for the test sets of the extension trip and competition trip models were 0.84 and 0.82, respectively, while the R2 value for the supplementary trip model was 0.76. [Conclusion] SHAP analysis further revealed the nonlinear impact of parking availability, proximity to the city center, and population density on the competitive and cooperative relationships between shared bikes and subway systems. The effect of parking availability first decreases and then increases, with extension travel remaining relatively stable, whereas supplementary travel follows an S-shaped curve, with diminishing marginal effects after reaching a certain threshold. Regarding proximity to the city center, the impact on shared-bike usage initially increased and then decreased. Specifically, the demand for extension travel was higher within 20 km, whereas competitive travel decreased with increasing distance, and the demand for supplementary travel diminished beyond 15 km. In terms of population density, there was an S-shaped growth in the demand for extension and supplementary travel, with a significant increase in shared bike demand in high-density areas. However, competitive travel was less influenced by the population density and depended more on the availability of subways and other public transportation options. [Application] The findings offer valuable insights for urban transportation planning. In areas with fewer parking facilities, increasing shared-bikes deployment can enhance short-distance transfer efficiency. Additionally, in areas within 20 km of the city center, strengthening the integration between shared bikes and subways can significantly improve the transfer experience. By optimizing the built environment, shared-bike usage can be increased, reliance on private cars reduced, traffic carbon emissions minimized, and a more sustainable and green transportation system can be achieved.
[1]YANG Y, HEPPENSTALL A, TURNER A, et al. A spatiotemporal and graph-based analysis of dockless bike sharing patterns to understand urban flows over the last mile[J]. Computers, Environment and Urban Systems, 2019,77:101361.
[2]郭源园,吴磊,曾鹏.都市区建成环境对“共享单车+地铁”通勤使用影响的空间异质性：以深圳市为例[J].热带地理, 2023, 43(5):872-884.GUO Yuanyuan, WU Lei, ZENG Peng. Spatial heterogeneity of the built environment effect on the use of a bikesharemetro commute in a metropolitan area:a case study of Shenzhen[J]. Tropical Geography, 2023, 43(5):872-884.
[3]YAN Y, CHEN Q. Spatial heterogeneity and nonlinearity study of bike-sharing to subway connections from the perspective of built environment[J]. Sustainable Cities and Society, 2024, 114:105766.
[4]YU J, CHANG X, HU S, et al. Combining travel behavior in metro passenger flow prediction:a smart explainable stacking-catboost algorithm[J]. Information Processing&Management, 2024, 61(4):103733.
[5]LV H, LI H, CHEN Y, et al. An origin-destination level analysis on the competitiveness of bike-sharing to underground using explainable machine learning[J]. Journal of Transport Geography, 2023, 113:103716.
[6]陈启香,吕斌,陈喜群,等.空间异质性建成环境对出租车与地铁竞合关系的影响[J].交通运输系统工程与信息, 2022, 22(3):25-35.CHEN Qixiang, LV Bin, CHEN Xiqun, et al. Impacts of built environment on competition and cooperation relationship between taxi and subway considering spatial heterogeneity[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(3):25-35.
[7]LINDSEY C, SIÂN M, ELIZABETH A W. The association between residential housing prices, bicycle infrastructure and ridership volumes[J]. Urban Studies, 2021, 58(4):787-808.
[8]LI A, ZHAO P, HUANG Y, et al. An empirical analysis of dockless bike-sharing utilization and its explanatory factors:case study from Shanghai, China[J]. Journal of Transport Geography, 2020, 88:102828.
[9]韦娇敏，刘卓，陈艳艳，等.考虑宏微观建成环境的共享单车骑行影响因素分析[J].科学技术与工程，2023, 23(9):3904-3915.WEI Jiaomin, LIU Zhuo, CHENYanyan, et al. Analysis of factors influencing dockless bike-sharing cycling considering macroscale and microscale build environment[J].Science Technology and Engineering, 2023, 23(9):3904-3915.
[10]LIU M, LIU Y, YE Y. Nonlinear effects of built environment features on metro ridership:an integrated exploration with machine learning considering spatial heterogeneity[J]. Sustainable Cities and Society, 2023, 95:104613.
[11]李聪颖,吴佳西,张洪涛,等.建成环境对城市机动车出行时空异质影响研究[J].交通运输工程与信息学报, 2024, 22(4):52-66.LI Congying, WU Jiaxi, ZHANG Hongtao, et al. Spatiotemporal heterogeneous effects of built environment on urban motorized travel[J]. Journal of Transportation Engineering and Information, 2024, 22(4):52-66.
[12]汪晴,崔珩.建成环境对共享单车高峰出行的非线性影响——以厦门主岛为例[J].南方建筑，2024(1):20-28.WANG Qing, CUI Heng. Nonlinear influences of built environment on travel with shared bicycles in rush hour:a case study of principal island in Xiamen[J]. South Architecture, 2024(1):20-28.
[13]李培坤,陈旭梅,鲁文博,等.基于XGBOOST-SHAP的地铁建成环境与站点出行距离的非线性关系研究[J].铁道科学与工程学报, 2024, 21(4):1624-1633.LI Peikun, CHEN Xumei, LU Wenbo, et al. Research on nonlinear relationship between subway built environment and travel distance of stations based on XGBOOST-SHAP[J]. Journal of Railway Science and Engineering, 2024, 21(4):1624-1633.
[14]HAN T, BIAN Y, ZHAO X. The application of XGBoost and SHAP to examine factors in bike sharing-related demand[C]//22th COTA International Conference of Transportation Professionals. Changsha:CICTP , 2022:2980-2990.
[15]FENG D C, WANG W J, MANGALATHU S, et al. Interpretable XGBoost-SHAP machine-learning model for shear strength prediction of squat RC walls[J]. Journal of Structural Engineering, 2021, 147(11):04021173.
[16]ZHANG Y, MENG Y, CHEN X J, et al. The nonlinear effect of the built environment on bike-metro transfer in different times and transfer flows considering spatial dependence[J]. Sustainability, 2025, 17(1):251.
[17]LU X, CHEN C, GAO R D, et al. Prediction of highspeed traffic flow around city based on BO-XGBoost model[J]. Symmetry, 2023, 15(7):1453.
[18]马良玉,於世磊,赵尚羽,等.基于随机搜索算法优化XGBoost的过热汽温预测模型[J].华北电力大学学报（自然科学版）, 2021, 48(4):99-105.MA Liangyu, YU Shilei, ZHAO Shangyu, et al. Superheated steam temperature prediction models based on XGBoost optimized with random search algorithm[J].Journal of North China Electric Power University(Natural Science Edition), 2021, 48(4):99-105.
[19]BERGSTRA J, BENGIO Y. Random search for hyperparameter optimization[J]. Journal of Machine Learning Research, 2012, 13:281-305.
[20]JI S, WANG X, LYU T, et al. Understanding cycling distance according to the prediction of the XGBoost and the interpretation of SHAP:a non-linear and interaction effect analysis[J]. Journal of Transport Geography, 2022,103:103414.
[21]CALTHORPE P. The next American metropolis:ecology, community, and the American dream[M]. New York:Princeton Architectural Press, 1993.
[22]LIU C, SHEN Q. An empirical analysis of the influence of urban form on household travel and energy consumption[J]. Computers, Environment and Urban Systems,2011, 35(5):347-357.
[23]刘柯良,陈坚,祝烨,等.社区建成环境对小汽车使用行为的非线性影响模型[J].北京交通大学学报, 2022,46(3):49-56.LIU Keliang, CHEN Jian, ZHU Ye, et al. Non-linear impact model of community built environment on car usage behavior[J]. Journal of Beijing Jiaotong University,2022, 46(3):49-56.
[24]NI Y, CHEN J. Exploring the effects of the built environment on two transfer modes for metros:dockless bike sharing and taxis[J]. Sustainability, 2020, 12(5):2034.
[25]唐秋生,许浩.建成环境非线性作用下汽车客运站最高聚集人数模型[J].重庆理工大学学报（自然科学）,2022, 36(6):216-223.TANG Qiusheng, XU Hao. Model of the maximum number of people gathered in a bus terminal considering the non-linear effect of the built environment[J]. Journal of Chongqing University of Technology(Natural Science),2022, 36(6):216-223.
[26]杨朝斌.城市空间结构对城市热环境时空异质性分布影响研究[J].测绘学报, 2022, 51(5):788-788.YANG Chaobin. The effect of the urban spatial structure on the spatio-temporal patterns of the urban thermal environment[J]. Acta Geodaetica et Cartographica Sinica,2022, 51(5):788-788.
[27]高楹,宋辞,舒华,等.北京市摩拜共享单车源汇时空特征分析及空间调度[J].地球信息科学学报, 2018, 20(8):1123-1138.GAO Ying, SONG Ci, SHU Hua, et al. Spatial-temporal characteristics of source and sink points of mobikes in Beijing and its scheduling strategy[J]. Journal of Geo-Information Science, 2018, 20(8):1123-1138.
[28]廖洋,罗霞,王红杰.考虑时空自相关性的共享电动汽车出行选择影响因素分析[J].交通运输工程与信息学报, 2024, 22(4):67-84.LIAO Yang, LUO Xia, WANG Hongjie. Analysis of factors influencing shared electric vehicle travel choice considering spatiotemporal autocorrelation[J]. Journal of Transportation Engineering and Information, 2024, 22(4):67-84.
[29]樊东卫,何勃亮,李长华,等.球面距离计算方法及精度比较[J].天文研究与技术, 2019, 16(1):69-76.FAN Dongwei, HE Boliang, LI Changhua, et al. Research on spherical distance computation and accuracy comparison[J]. Astronomical Research&Technology,2019, 16(1):69-76.
[30]黎子豪,蒋恕.基于机器学习和SHAP算法的声波测井曲线重构及可解释性分析[J].地质科技通报, 2025, 44(1):321-331.LI Zihao,JIANG Shu. Reconstructing and interpreting analysis of sonic logging curves based on machine learning and SHAP algorithm[J]. Bulletin of Geological Science and Technology,2025,44(1):321-331.
[31]李浩,卢朝阳,谈翌平,等.基于MIC-iAFF-Stacking集成学习的航空器滑出时间预测[J].交通运输工程与信息学报, 2024, 22(4):142-153.LI Hao, LU Chaoyang, TAN Yiping, et al. Aircraft departure taxi time prediction based on MIC-iAFF-Stacking ensemble learning[J]. Journal of Transportation Engineering and Information, 2024, 22(4):142-153.
[32]刘琪琪，陈春，匡新晖.利用XGBoost模型查明土地利用格局对行人交通事故严重程度的非线性影响[J].科学技术与工程，2025, 25(3):1253-1261.LIU Qiqi, CHEN Chun, KUANG Xinhui. Identifying non-lineareffects of land use patterns on pedestrian crash severity using XGBoost decisiontreen model[J]. Science Technology and Engineering, 2025, 25(3):1253-1261.
[33]杨逍遥,王建军,李国栋,等.基于GWR模型的共享单车空间分布影响因素分析[J].交通运输研究, 2021, 7(1):81-94.YANG Xiaoyao, WANG Jianjun, LI Guodong, et al. Factors impacting spatial distribution of bike sharing based on GWR model[J]. Transport Research, 2021, 7(1):81-94.
[34]李杰杰,孙昌盛,张春英,等.基于多尺度地理加权回归模型的城市住宅价格影响因素空间异质性:以桂林市主城区为例[J].浙江大学学报（理学版）, 2024, 51(5):623-635.LI Jiejie, SUN Changsheng, ZHANG Chunying, et al.Spatial heterogeneity of influential factors of urban housing prices based on multiscale geographically weighted regression models:a case study of the main urban area of Guilin city[J]. Journal of Zhejiang University(Science Edition), 2024, 51(5):623-635.
[35]陈启香,吕斌,李显林.站域建成环境对出租车-地铁组合出行的非线性影响[J].交通运输工程学报, 2024, 24(5):285-300.CHEN Qixiang, LYU Bin, LI Xianlin. Nonlinear effect of station-area built environment on taxi-metro combined travel[J]. Journal of Traffic and Transportation Engineering, 2024, 24(5):285-300.
(1)OpenStreetMap(OSM)是一个由全球志愿者共同协作创建的免费开源的地图平台（https://www.openstreetmap.org），每个人都可在该平台上获取可编辑的地图数据，并且可以自由使用这些数据进行创作、分享以及改进，从而促进地理信息的开放与共享。
(1)LandScan官网是访问全球人口分布数据库的主要入口（https://landscan.ornl.gov/），是由美国橡树岭国家实验室开发的一个高分辨率全球人口动态数据集。该数据库提供了详细的人口分布信息，空间分辨率为30弧秒（约1 km2），覆盖了全球范围。
基本信息:
DOI：10.19961/j.cnki.1672-4747.2025.02.001
中图分类号:U12;F572;F724.6
引用信息:
[1]马健兵,张伊涵,张永琪.建成环境视角下共享单车与地铁的竞合关系非线性影响分析[J].交通运输工程与信息学报,2026,24(02):22-37.DOI:10.19961/j.cnki.1672-4747.2025.02.001.
基金信息:
四川省重点研发计划资助项目（2023YFS0192）
2025-04-14
2025-04-14
2025-04-14