原文传递
Analysis of Spatiotemporal Factors Affecting Traffic Safety Based on Multisource Data Fusion
| 题名： | Analysis of Spatiotemporal Factors Affecting Traffic Safety Based on Multisource Data Fusion |
| 正文语种： | eng |
| 作者： | Cheng Xu;Zuoming Zhang;Fengjie Fu;Wenbin Yao;Hongyang Su;Youwei Hu;Donglei Rong;Sheng Jin |
| 作者单位： | Dept. of Traffic Management Engineering Zhejiang Police College No. 555 Binwen Rd. Hangzhou 310053 China;Polytechnic Institute Zhejiang Univ. No. 269 Shixiang Rd. Hangzhou 310015 China;Dept. of Traffic Management Engineering Zhejiang Police College No. 555 Binwen Rd. Hangzhou 310053 China;Institute of Intelligent Transportation Systems Zhejiang Univ. No. 866 Yuhangtang Rd. Hangzhou 310058 China;Polytechnic Institute Zhejiang Univ. No. 269 Shixiang Rd. Hangzhou 310015 China;Zhejiang Lab Zhongtai St. Kechuang Ave. Hangzhou 311121 China;Institute of Intelligent Transportation Systems Zhejiang Univ. No. 866 Yuhangtang Rd. Hangzhou 310058 China;Institute of Intelligent Transportation Systems Zhejiang Univ. No. 866 Yuhangtang Rd. Hangzhou 310058 China |
| 关键词： | Traffic safety; Traffic state; Road network structure; Built environment; Geographical and temporal weighted regression (GTWR) model; Complex network |
| 摘要： | Traffic state information, road network structure characteristics, and built environment characteristics are factors influencing traffic safety, which will alleviate or aggravate traffic safety problems. This paper analyzes the relationship between these factors and traffic accidents involving either property damage only (PDO) crashes or killed and severe injury (KSI) crashes. The spatiotemporal distribution of the two types of accidents was analyzed. Abundant explanatory variables were extracted from accident data, license plate recognition (LPR) data, OpenStreetMap (OSM) data, and point of interest (POI) data based on complex network methods and information entropy theories. Geographical and temporal weighted regression (GTWR), geographically weighted regression (GWR), and ordinary least squares (OLS) models were used to analyze the influencing factors of traffic safety, respectively. The results demonstrate that the GTWR model performs best in modeling spatiotemporal heterogeneity data. Traffic state information, road network structure, and built environment factors all have significant effects on traffic accidents, and traffic state information have the highest correlation with traffic accidents among all factors. The greater the traffic volume, the more likely are traffic accidents. The strongest correlation is between PDO crashes and traffic state in the morning peak, in the evening peak, and at night. For a road network divided into grids, the more important the intersections in the grid, the greater is the street circuity, and the more chaotic the street direction, the more likely PDO crashes are to occur in the grid. Furthermore, the diversity of land use is positively correlated with traffic accidents in urban areas, whereas the correlation is negative in suburban areas, which reflects the spatial heterogeneity. |
| 出版年： | 2023 |
| 期刊名称： | Journal of Transportation Engineering |
| 卷： | 149 |
| 期： | 10 |
| 页码： | 04023098.1-04023098.12 |
相关文献
- Expressway Traffic Incident Detection Using a Deep Learning Approach Based on Spatiotemporal Features with Multilevel Fusion
- Traffic Order Analysis of Intersection Entrance Based on Aggressive Driving Behavior Data Using CatBoost and SHAP
- Exploring Affecting Factors of Park Use Based on Multisource Big Data: Case Study in Wuhan, China
- Establishing Multisource Data-lntegration Framework for Transportation Data Analytics
- Freeway Traffic State Estimation Method Based on Multisource Data
- Traffic Propagation in Road Network from a Data-Driven Analysis Perspective
检索历史
应用推荐
