Document Type
Article
Publication Date
7-2017
Publication Title
IEEE Transactions on Intelligent Transportation System
Volume
18
Issue
7
Publisher
IEEE
DOI
https://doi.org/10.1109/TITS.2017.2658664
Abstract
A novel framework is developed in this paper, to increase the real-time roadway traffic condition assessment accuracy, which integrates connected vehicle technology (CVT) with artificial intelligence (AI) paradigm forming a CVT-AI method. Traffic density is a major indicator of traffic conditions. In this paper, the traffic operational condition is assessed based on traffic density. A simulated network of Interstate 26 in South Carolina is developed to investigate the effectiveness of the method. The assumption is that the vehicle onboard units will forward the CV generated data to the edge devices (e.g., roadside units) for further processing. CV generated distance headway and number of stops, and speed data are used to estimate traffic density. This paper reveals that, with 20% and greater CV penetration levels, the accuracy of the density information with the AI-aided CVT is a minimum of 85%. Moreover, this paper demonstrates that the integrated CVT-AI method yields a higher accuracy with the increase of CV penetration levels. Level of service (LOS) is the indicator of traffic congestion level on highways and is described with traffic density in terms of passenger car/mile/lane for a specific free flow speed. LOS estimated using the CVT-AI density estimation method is compared with the density estimation algorithm used by the Caltrans Performance Measurement System (PeMS), which relies on the occupancy and flow data collected by the freeway inductive loop detectors. With a 10% or more CV penetration, higher accuracy is achieved using the CVT-AI algorithm compared with the PeMS density estimation algorithm.
Recommended Citation
S. M. Khan, K. C. Dey and M. Chowdhury, "Real-Time Traffic State Estimation With Connected Vehicles," in IEEE Transactions on Intelligent Transportation Systems, vol. 18, no. 7, pp. 1687-1699, July 2017, doi: 10.1109/TITS.2017.2658664.
Comments
The published version of this article can be found here: https://ieeexplore.ieee.org/document/7878538