Spatiotemporal variations of private e-bike trips with explainable data-driven technologies
文献类型：期刊论文
| 作者 | Wang, Peixiao3,4; Zhang, Hengcai3,4; Zhang, Beibei3,4; Cheng, Shifen3,4; Lu, Feng2,3,4; Zhang, Tong1 |
| 刊名 | CITIES
 |
| 出版日期 | 2025-03-01 |
| 卷号 | 158页码:105712 |
| 关键词 | Private e -bike trips Trip patterns Spatiotemporal variations Driving mechanisms Spatiotemporal random forest Improved SHAP |
| ISSN号 | 0264-2751 |
| DOI | 10.1016/j.cities.2025.105712 |
| 产权排序 | 1 |
| 文献子类 | Article |
| 英文摘要 | Understanding the trip features and driving mechanisms of e-bikes, particularly their spatiotemporal variations, is essential for improving traffic mobility, reducing pollution, and enhancing road safety. Currently, existing studies have two main gaps: (1) the spatiotemporal variations of private e-bikes remain unclear, and (2) there is a lack of explainable data-driven techniques that can analyze the spatiotemporal variation effects of driving mechanisms, especially considering spatiotemporal heterogeneity. In this study, using the private e-bikes trips in Wuhan, China as a case study, a novel explainable framework is proposed to analyze the spatiotemporal variations in their trip features and driving mechanisms. More specifically, a novel spatiotemporal random forest is presented to build a nonlinear mapping between driving factors and private e-bike trips in the spatiotemporal domain. Then, the classical SHAP method is extended to map Shapley values onto the time and space axes, enabling the exploration of spatiotemporal variations in driving factors. Findings reveal that: (1) private e-bikes are frequently used for short and medium-distance trips, typically exceeding 1 km, and play a crucial role in daily urban commuting; (2) Factors such as Historical trip frequency, Commercial POI Density, and Hospital POI Density are positively correlated with private e-bike trips; (3) the influence of driving factors on private e-bike trips vary significantly across different spatial locations and time windows. This study offers an innovative analytical framework for a more profound comprehension of e-bike trips. Additionally, the findings can aid authorities in crafting more effective policies and planning strategies. |
| URL标识 | 查看原文 |
| WOS关键词 | TEMPORALLY WEIGHTED REGRESSION |
| WOS研究方向 | Urban Studies |
| 语种 | 英语 |
| WOS记录号 | WOS:001407813100001 |
| 出版者 | ELSEVIER SCI LTD |
| 源URL | [http://ir.igsnrr.ac.cn/handle/311030/212312]  |
| 专题 | 资源与环境信息系统国家重点实验室_外文论文 |
| 通讯作者 | Zhang, Hengcai |
| 作者单位 | 1.Wuhan Univ, State Key Lab Informat Engn Surveying Mapping & Re, Wuhan 430079, Peoples R China 2.Fujian Collaborat Innovat Ctr Big Data Applicat Go, Fuzhou 350003, Peoples R China; 3.Univ Chinese Acad Sci, Coll Resources & Environm, Beijing 100049, Peoples R China; 4.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, State Key Lab Resources & Environm Informat Syst, Beijing 100101, Peoples R China; |
| 推荐引用方式 GB/T 7714 | Wang, Peixiao,Zhang, Hengcai,Zhang, Beibei,et al. Spatiotemporal variations of private e-bike trips with explainable data-driven technologies[J]. CITIES,2025,158:105712. |
| APA | Wang, Peixiao,Zhang, Hengcai,Zhang, Beibei,Cheng, Shifen,Lu, Feng,&Zhang, Tong.(2025).Spatiotemporal variations of private e-bike trips with explainable data-driven technologies.CITIES,158,105712. |
| MLA | Wang, Peixiao,et al."Spatiotemporal variations of private e-bike trips with explainable data-driven technologies".CITIES 158(2025):105712. |
入库方式： OAI收割
来源：地理科学与资源研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。