Cooperative ACC uses a combination of forward ranging sensor(s) and vehicle-vehicle (V2V) communication to help a vehicle adjust its speed to follow the preceding vehicle in the same lane. The addition of the V2V communication enables the system to respond more quickly to speed changes by the preceding vehicle and vehicles ahead of it, beyond the line of sight. This greatly improves the stability of traffic flow, increases driver confidence in the system and enables shorter following distances. These, in turn, lead to energy savings and increases in the effective capacity of a highway lane. More broadly, CACC enables cooperative maneuvering by vehicles, facilitating merging of traffic streams and formation of vehicles into platoons. The proposed project will overcome the key remaining technical challenges that stand in the way of implementation of CACC, so that it can move closer to deployment and realization of the high-performance vehicle streams sought in the BAA.
For ACC/CACC, higher speeds will lead to longer following distances, somewhat similar to the behavior of common drivers. The ACC may be conservative or aggressive depending on the time gap setting. It is not exactly the same as a driver, however because (autonomous rather than cooperative) ACC only depends on remote sensor measurements relative to its immediate predecessor vehicle and cannot look/predict the behavior of the vehicles further forward, while an experienced driver usually looks/predicts the behaviors of several vehicles ahead. The limitation of autonomous ACC performance is the cumulative delays from the leader to the upstream of the vehicle-following string. This is the main destroyer of the string stability in vehicle following, which is the reason why higher ACC market penetration will makes traffic stability worse. CACC, on the other hand, provides prediction capability through Vehicle-to-Vehicle (V2V) communication: vehicle onboard sensor data including CAN (Control Area Network) Bus data are passed to all the vehicles behind, which is more accurate and with less time delay than driver perception. This is the reason why CACC vehicles can provide more stable car following than experienced drivers. In addition, significant differences among the drivers are eliminated.
The technical approach in the project is designed to capitalize on the extensive prior knowledge and experience of the project team with CACC, cooperative ITS in general, driver behavior, and traffic dynamics modeling. The project begins with a careful identification of the operational concept alternatives that should be treated in more depth, so that a full range of reasonable alternatives is considered. These alternatives will include special infrastructure treatments such as managed lanes with differing degrees of separation from the mixed flow lanes and multiple strategies for assembling and disassembling CACC platoons. The first key milestone will be the documentation of these alternatives within four months of the project start.
Because experimental evaluation of the larger-scale maneuvering strategies would require large numbers of vehicles, the preferred approach for assessing these system management issues is in microsimulation. Therefore, the balance of the first year of effort is devoted to selection of suitable deployment scenarios to use for evaluation and to development of the simulation models that will represent the cooperative maneuvering strategies. The deployment scenarios will be chosen to represent diverse conditions (urban vs. rural, different O-D patterns and traffic volumes, and highways as well as signalized arterials) so that the maneuvering strategies can be matched to the situations in which they work most effectively. The development of the simulation models will be divided between the PATH and Delft team members, who will share these models for implementation on their respective simulation platforms (Aimsun and MOTUS). Each team will test the modules developed by the other team so that they will all be independently vetted by experts of high international standing, providing the results with high credibility in the research community. The simulation model development completion and validation milestone will be at the end of the first year.
The second year of the project is primarily devoted to use of the simulation tools to assess the traffic flow and energy impacts of the various maneuvering strategies under the different deployment scenarios, leading to a comprehensive report on those impacts by the ninth month of the second year. In parallel with that work, the experimental data that PATH has collected in tests of its two generations of CACC systems will be analyzed to reveal some important aspects of system performance and driver behavior that have not been studied previously. This will help expand understanding of how drivers choose to use both ACC and CACC and of the safety implications of the shorter gaps enabled by CACC, for a report to be completed by the end of the second year. Based on the results of the simulation studies, the project team will also identify the next generation of vehicle experiments that should be conducted to verify the technical feasibility and driver acceptability of the maneuvering strategies that appear to be most promising.
- X. Y. Lu, C. Nowakowski, and S.E. Shladover, et al, Partial Automation for Truck Platooning, presentation at Automated Vehicle Symposium, Ann Arbor, Michigan, July 20-24, 2015, PDF
- X. Y. Lu, S.E. Shladover, C. Nowakowski, Dali Wei, and R. Ferlis, Using Cooperative ACC to Form High-Performance Vehicle Streams, presented at Automated Vehicle Symposium, Ann Arbor, Michigan, July 20-24, 2015, PDF
- X. Y. Lu and S. Shladover, Automated Truck Platoon Control and Field Test, Road Vehicle Automation, Editors: Gereon Meyer and Sven Beiker, Lecture Notes in Mobility, ISSN: 2196-5544, Springer, 2014