Real-time estimation of multi-class path travel times using multi-source traffic data

作者：Li, Ang^*; Lam, William H. K.; Ma, Wei; Wong, S. C.; Chow, Andy H. F.; Tam, Mei Lam

来源：EXPERT SYSTEMS WITH APPLICATIONS, 2024, 237: 121613.

DOI：10.1016/j.eswa.2023.121613

• Real -time travel time estimation
• Multiple vehicle classes
• AVI data
• GPS data
• Point sensor data

In practice, most of the intelligent transportation systems provide average travel times of all vehicles on selected paths in real time on a regular basis. However, path travel times of different vehicles could vary widely under different traffic conditions. There is a need to consider the differences in vehicle classes for path travel time estimation. This paper proposes a novel modeling framework that considers variance-covariance relationships between vehicle classes for real-time estimation of multi-class path travel times with use of multi-source traffic data collected from various types of sensors. The proposed methodology is examined with a case study of a selected urban expressway in Hong Kong with data obtained from multiple sources. The path travel time esti-mates by vehicle class are validated and the results demonstrate the merits and performance of the proposed framework.