Connected and Automated Vehicles (CAVs) Leveraged to Enhance Efficiency

Lead Investigator: 
Andreas Malikopoulos
Participating Staff: 
Wei Lu
Jason Carter
Collaborators: 
Boston University and University of Virginia
Sponsors: 
LDRD Program at ORNL
Start Date: 
2014
End Date: 
2016

The goals of the next generation transportation systems are to alleviate congestion, reduce energy use and emissions, and improve safety. Core disruptive technologies for such transportation systems include vehicle connectivity, vehicle automation, and the notion of shared personalized infrastructure enabled by mobility on demand systems. The central challenge is to develop more efficient transportation systems to connect communities and increase access, without also increasing the negative consequences of transportation (e.g., emissions, energy consumption, and congestion). Connected and automated vehicles (CAVs) provide the most intriguing opportunity for enabling users (including individual vehicles and traffic control centers) to better monitor transportation network conditions and make better operating decisions to improve safety and reduce pollution, energy consumption, and travel delays.

 

 

Significance

Recognition of the necessity of CAVs is gaining momentum. Many stakeholders intuitively see the benefits of multiscale vehicle control systems and have started to develop business cases for their respective domains, including the automotive and insurance industries, government and service providers. It seems clear that the availability of vehicle-to-vehicle communication has the potential to reduce traffic accidents and ease congestion by enabling vehicles to more rapidly account for changes in their mutual environment. Likewise, vehicle-to-infrastructure communication, e.g., communication with traffic structures, nearby buildings, and traffic lights, should allow for individual vehicle control systems to account for unpredictable changes in local infrastructure.

 

The overarching goal of this project is to establish a rigorous optimization framework and develop decentralized control algorithms for online coordination of CAVs. The underlying concept hinges on the idea that in the “new world” of massive amounts of data from vehicles and infrastructure, what we used to model as uncertainty (noise or disturbance) becomes additional input or extra state information in a much higher-dimensional vector. The vehicle can no longer operate in isolation and must integrate external data and being considered as part of the entire system. The direct implication of this approach is that it changes the mathematical framework of how traffic flow is optimized by converting noise-based models into almost entirely deterministic ones with larger state spaces. In this context, we have produced the following outcome:

 

  • Modeled mathematically vehicle interactions in different transportation segments, e.g., intersections, merging roadways, speed harmonization in highways.
  • Formulated and solved an optimal control problem, the solution of which yields the optimal coordination of CAVs in these transportation segments.
  • Quantified the impact of these concepts in terms of fuel consumption and travel time, where it was shown that fuel consumption can be improved by up to 65% whereas travel time by up to 35%.