Integrated Traffic Flow Control in a Connected Network
The purpose of this project is to analyze the dynamics of traffic flow especially during incidents and bottlenecks where these dynamics are excited and create the congestion phenomena we observe in practice, and come up with techniques to better control traffic flow. While the techniques we plan to develop can be applied in todays system, connectivity such as vehicle to infrastructure communication and/or vehicle to vehicle communication will allow the realization of their full benefits. Building on very promising preliminary results we plan to develop lane change and variable speed controllers which will integrate with ramp metering control and traffic light controllers in order to better control traffic flow especially during incidents and bottlenecks. Our preliminary results are based on treating all lanes in a highway as having the same density and traffic flow. In this project, we plan to model the flow in each lane separately and allow different flow characteristics in each lane. This model is more appropriate for highway lanes where the volume of trucks is relatively high and tends to concentrate on the slow lanes. We plan to use the cell transmission model and triangular fundamental diagram, that was proven adequate in our previous studies for control design purposes, to model the flow in each lane. We will then use the model to develop variable speed limit controllers which will integrate with ramp metering and lane change controllers. We plan to use a spatial model for developing lane change controllers using an optimization procedure rather than the preliminary adhoc method we used in previous studies. The lane change controller will generate commands for drivers to change lanes before reaching the incident or bottleneck in order to maximize throughput and reduce congestion.
Funding source: USDOT
Funding amount: $100,000
Start and end dates: 1/1/2018 to 12/31/2018
The data associated with this report are available at: https://doi.org/10.7910/DVN/SPOKSM