A Three-Dimensional Spatiotemporal Model Concepts and Its Application on Traffic Flow Analysis Using Trajectory Data

• Published in: 2019 IEEE Intelligent Transportation Systems Conference (ITSC) pp. 399 - 406
• Main Authors: Che, Xueqi, Wei, Yanning, Li, Keping, Pan, Maolin
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2019
• Subjects: Analytical models; Data models; Roads; Safety; Solid modeling; Three-dimensional displays; Trajectory
• DOI: 10.1109/ITSC.2019.8917312

The characteristics of traffic flow at intersections are critical to traffic safety and efficiency analysis. In the context of big data, massive trajectory data becomes an important source for studying traffic flow. Currently, the two-dimensional time-space diagram is the primary visualization tool used to represent trajectory data. Most existing models mainly focus on the longitudinal behaviors while ignoring the lateral movement, which has limitations in representing realistic interactions among vehicles. This paper proposes a three-dimensional (3D) spatiotemporal analysis model using trajectory data. By establishing a complete space-time dimension, it is possible to simultaneously study traffic flow characteristics and vehicle conflicts at intersections in the visual 3D model. First, the basic definitions and assumptions of the 3D coordinate system are defined. The analysis function is then illustrated from different perspectives of the 3D model. Finally, two concepts and their calculation methods are introduced, which are the space-time utilization for efficiency analysis and the shortest space-time distance for traffic safety evaluation. The proposed model is applied with the simulated trajectory data of typical signalized intersections. In conclusion, the developed 3D analysis model provides a convenient and comprehensive visualization tool for traffic flow analysis, both in terms of safety evaluation and efficiency analysis.