Inefficient mobility management leads to congestion with several socioeconomic and environmental adverse effects. Despite numerous research efforts and deployed technological solutions to address congestion, the problem still persists. One of the main reasons is attributed to the sparsity and low quality of traffic data. Recent advances in information and communication technologies have led to the emergence of new sensing devices with numerous capabilities for the collection of traffic data, that introduce new opportunities to traffic monitoring. This work provides an overview of such emerging sensing technologies and focuses on fine-grained traffic density estimation. A general streaming-data-driven probabilistic approach is designed to utilize different types of measurements. In this work, we examine how to utilize measurements from (i) fixed-location sensors, (ii) Connected and Automated Vehicles (CAVs), and (iii) Unmanned Aerial Vehicles (UAVs). The proposed methodology integrates the Gaussian process model within a Bayesian framework to effectively calculate traffic density estimates for different measurement types, even in cases where only a limited number of measurements are available within the designated time–space area being examined. This is achieved by using the available information of the time–space vehicle trajectory diagram that yields traffic density measurements which are extrapolated in time and space where no information has been obtained. A simulation study is performed to evaluate the estimation methodology utilizing measurements from the three different sensing technologies. Estimation results show that each sensing technology can be effective under different settings with their main difference being that CAVs and UAVs offer information with higher spatiotemporal resolutions compared to fixed-location sensors, while UAV-based sensing is effective even under low penetration rates of UAVs flying above the network and low percentages of network coverage.
Streaming-Data-Driven Traffic Density Estimation Using Gaussian Processes
Springer Tracts on Transportation, Traffic
2024-12-22
34 pages
Article/Chapter (Book)
Electronic Resource
English
Traffic State Estimation from Vehicle Trajectories with Anisotropic Gaussian Processes
ArXiv | 2023
|Traffic density estimation using vehicle sensor data
Taylor & Francis Verlag | 2022
|