Field Implementation Feasibility Study of Cumulative Travel-Time Responsive Intersection Control Algorithm under Connected Vehicle Technology
This project utilized the Connected Vehicle (CV) environment, which provides two-way wireless communications between vehicles and infrastructure, to (1) improve the Cumulative Travel-time Responsive (CTR) Intersection Control Algorithm under low CV market penetration by utilizing Bluetooth technology, and (2) assess potential benefits of the CTR algorithm by examining mobility, energy, and greenhouse emissions measures. The project team developed and evaluated a hardware-in-the-loop simulation to ensure that the developed CTR algorithm will work with an existing traffic controller on the Northern Virginia Connected Vehicle Test Bed. The team enhanced the CTR algorithm and evaluated its impact to verify the feasibility of field implementation. Two prediction techniques, a standard Kalman filter (SKF) and an adaptive Kalman filter (AKF), were applied to estimate cumulative travel time for each phase in the CTR algorithm. In addition, traffic demand, the market penetration rate (MPR), and the types of available data were also considered in evaluating CTR algorithm performance. The Lee Highway and Nutley Street intersection on the Northern Virginia Connected Vehicle Test Bed was selected for a case study and simulated within VISSIM. The results showed that the CTR algorithm’s performance depends on the MPR, as the information collected from CVs is a key CTR algorithm-enabling factor. However, this study found that the MPR could be relaxed (1) when the data were collected from both CV and infrastructure sensors, and (2) when an AKF was adopted in the CTR algorithm. The minimum MPRs required to outperform the current actuated traffic signal control were empirically found for each prediction technique and types of available data—data from both Connected Vehicle and infrastructure sensors, or Connected Vehicle’s data only. Even without the infrastructure sensors, the CTR algorithm could be considered for implementation at an intersection with high traffic demand and a 50% to 60% MPR. As the MPR for this field evaluation was around 14%, much lower than the minimum 20% required with an AKF incorporated, the project team could not implement the proposed algorithm. Instead, the team developed an implementation plan that can be easily adopted by traffic engineers once the MPR reaches 20% or higher.