Virginia Tech Transportation Institute

ARPA-E Traveler Response Architecture using Novel Signaling for Network Efficiency in Transportation (TRANSNET) Project

The Center for Sustainable Mobility teamed with Xerox' Palo Alto Research Center (PARC) that entailed developing the Collaborative Optimization and Planning Transportation Energy Reduction (COPTER) controller---a complete solution for the TRANSNET goal with comprehensive transportation network modeling, a decision-theoretic approach for system optimization, and explicit human behavior and influence modeling to maximize real-world impact. This project leveraged PARC's competencies in model based control of complex systems and human cognitive modeling, CSM's recognized leadership in transportation modeling and control, and Xerox's substantial incumbency as a provider of transportation service solutions to US cities to create a project that was meaningful, executable, and transitionable. CSM build a large-scale agent-based multi-modal model (cars, buses, trains, bicycles and walking) that was used to model over 3 million travelers in the Greater Los Angeles Area over the morning and evening peak periods. The project concluded that with 10% of the Greater Los Angeles peak period population receiving COPTER messages that 55% of those receiving the messages (i.e. 5.5% of the peak travelers) would accept these recommendations and this would produce up to a 4% reduction in energy and 20% reduction in delay over the entire Greater Los Angeles Area.

Developing an Eco-Cooperative Automated Control System (Eco-CAC)

This work is being funded by the Department of Energy Office of Energy Efficiency and Renewable Energy (EERE), Vehicle Technologies Office, Energy Efficient Mobility Systems Program under award number DE-EE0008209. Given that the transportation sector accounts for 69% of the nation’s petroleum consumption and 33% of the nation’s CO2 emissions, any reductions in the energy consumed by the transportation sector will have significant environmental benefits. Connected Vehicle (CV) systems comprise sets of applications that connect vehicles to each other and to the roadway infrastructure using vehicle-to-vehicle (V2V) and vehicle-to-infrastructure V2I communications, collectively known as V2X. While Automated Vehicles (AVs) enhance the operation of individual vehicles, CVs produce cooperative, network-wide benefits through the exchange of information. These new technological advancements have the potential to drastically improve the efficiency and sustainability of our transportation system. The main project objective is to substantially reduce vehicle fuel/energy consumption by integrating vehicle control strategies with CAV applications. Specifically, we are developing a novel integrated control system that (1) routes vehicles in a fuel/energy-efficient manner and balances the flow of traffic entering congested regions, (2) selects vehicle speeds based on anticipated traffic network evolution to avoid or delay the breakdown of a sub-region, (3) minimizes local fluctuations in vehicle speeds (also known as speed volatility), and (4) enhances the fuel/energy efficiency of various types of vehicles while focusing on internal combustion engine vehicles (ICEVs), battery electric vehicles (BEVs), hybrid electric vehicles (HEVs), and plug-in hybrid electric vehicles (PHEVs).

Developing Collaborative Connected Automated Vehicle Lane Selection and Merging Algorithms

This effort is being conducted in collaboration with Toyota and as part of the University Mobility Equity Center (UMEC). The research entails developing algorithms to identify the optimum lane allocation of CAVs along freeways and developing collaborative lane changing algorithms using game theory. It is anticipated that this research will improve the flow of traffic along freeway sections.

Traffic Signal Control within a Connected Vehicle Environment

This on-going effort entails developing real-time traffic signal control algorithms that use CV data. The effort has resulted in the development of a de-centralized cycle-free Nash bargaining traffic signal controller. The controller is currently being tested on different networks and has produced very promising results.

Multi-modal Network-wide Traffic State Prediction

The objective of this collaborative research effort with Ford is to develop a multi-modal trip departure and routing system considering a desired time of arrival at one’s destination while accounting for traveler response to these recommendations.