Behaviorally Compatible, Energy Efficient, and Network-Aware Vehicle Platooning Using Connected Vehicle Technology
pdf
By linking the transportation network to the communication grid, connected vehicle (CV) technology facilitates communication among vehicles, the infrastructure, and other road users. This connectivity is enabled through the Dedicated Short Range Communication (DSRC) technology, and creates a synergistic effect that paints a more comprehensive picture of the transportation network than could be observed by each individual road user. As such, it is expected that upon deployment, CV technology would significantly improve mobility, enhance traffic flow stability, reduce congestion, increase energy efficiency, and improve fuel economy, among other benefits.
By enabling communication, CV technology allows for creating cooperative automated driving systems that can lead to more effective coordination of the traffic stream. Such cooperative systems are manifested in form of platoons. A platoon is a single-file line (i.e., a virtual train) of vehicles which, owing to constant communication, are able to travel with small gaps between them. Platoon formation can introduce many benefits including (i) creating energy efficiency through reducing the aerodynamic drag force on platoon members; (ii) increasing road capacity through reducing the following distances between vehicles; (iii) reducing stochasticity in the traffic stream through having platoon members follow the lead of the platoon head, thereby reducing traffic congestion, improving travel times, and increasing travel time reliability; and (iv) facilitating real-time management of traffic and improving mobility by aggregating the unit of traffic from an individual vehicle to a cluster of vehicles.
Previous theoretical and experimental analysis conducted for automated platooning have revealed two important facts: (i) Platoons “are not all born equal”, and one platoon formation may be “greener” than another; and (ii) Position matters. More specifically, the reduction of fuel consumption distributes unevenly among vehicles in a platoon. Depending on the platoon formation and composition, each vehicle in the platoon may benefit differently from platooning, as the aerodynamic drag effects differ at each position. The highest drag saving normally occurs for middle vehicles while the lead vehicle has the least amount of drag reduction. It is easy to see that a vehicle may maximize its individual utility by frequently switching between platoons, as the vehicle’s immediate neighborhood evolves during a trip, creating instability in the traffic stream. Hence, In order for such benefits to be realized, platoons should have “stable” structures. In a stable platoon structure there does not exist any coalition of vehicles who would all increase their utilities by trading their platoon memberships.
Although an RSU allows for forming locally stable platoons, failing to account for network-level conditions could still lead to instability in the traffic stream. Examples include when a platoon has to dissolve when traveling through an off-ramp, or when new vehicles join the stream through an on-ramp, on a freeway stretch. Having vehicles leave platoons or new vehicles join the system could make the formerly stable platoons unstable; hence, network-level information need to be incorporated into the decision making process. CV technology can provide granular data at the network level by strategically positioning RSUs to ensure connectivity throughout the entire network. However, processing and making decisions using such high volume of data cannot occur in real-time. Therefore, we incorporate the network-level data at an aggregate level, and within a probabilistic framework, in the trajectory planning process.
Not only has the CV technology enabled platooning, but it has provided the ground to allow for efficient, effective, and robust vehicle-to-platoon assignments as well as real-time management of platoons using both local and network-level information. This proposal introduces a holistic framework that allows for (i) finding stable platoon structures, and (ii) planning trajectories that would allow for forming such platoons, based on both fine-level local and coarse-level network-wide information.
Submitted by Yafeng Yin
on
By linking the transportation network to the communication grid, connected vehicle (CV) technology facilitates communication among vehicles, the infrastructure, and other road users. This connectivity is enabled through the Dedicated Short Range Communication (DSRC) technology, and creates a synergistic effect that paints a more comprehensive picture of the transportation network than could be observed by each individual road user. As such, it is expected that upon deployment, CV technology would significantly improve mobility, enhance traffic flow stability, reduce congestion, increase energy efficiency, and improve fuel economy, among other benefits.
By enabling communication, CV technology allows for creating cooperative automated driving systems that can lead to more effective coordination of the traffic stream. Such cooperative systems are manifested in form of platoons. A platoon is a single-file line (i.e., a virtual train) of vehicles which, owing to constant communication, are able to travel with small gaps between them. Platoon formation can introduce many benefits including (i) creating energy efficiency through reducing the aerodynamic drag force on platoon members; (ii) increasing road capacity through reducing the following distances between vehicles; (iii) reducing stochasticity in the traffic stream through having platoon members follow the lead of the platoon head, thereby reducing traffic congestion, improving travel times, and increasing travel time reliability; and (iv) facilitating real-time management of traffic and improving mobility by aggregating the unit of traffic from an individual vehicle to a cluster of vehicles.
Previous theoretical and experimental analysis conducted for automated platooning have revealed two important facts: (i) Platoons “are not all born equal”, and one platoon formation may be “greener” than another; and (ii) Position matters. More specifically, the reduction of fuel consumption distributes unevenly among vehicles in a platoon. Depending on the platoon formation and composition, each vehicle in the platoon may benefit differently from platooning, as the aerodynamic drag effects differ at each position. The highest drag saving normally occurs for middle vehicles while the lead vehicle has the least amount of drag reduction. It is easy to see that a vehicle may maximize its individual utility by frequently switching between platoons, as the vehicle’s immediate neighborhood evolves during a trip, creating instability in the traffic stream. Hence, In order for such benefits to be realized, platoons should have “stable” structures. In a stable platoon structure there does not exist any coalition of vehicles who would all increase their utilities by trading their platoon memberships.
Although an RSU allows for forming locally stable platoons, failing to account for network-level conditions could still lead to instability in the traffic stream. Examples include when a platoon has to dissolve when traveling through an off-ramp, or when new vehicles join the stream through an on-ramp, on a freeway stretch. Having vehicles leave platoons or new vehicles join the system could make the formerly stable platoons unstable; hence, network-level information need to be incorporated into the decision making process. CV technology can provide granular data at the network level by strategically positioning RSUs to ensure connectivity throughout the entire network. However, processing and making decisions using such high volume of data cannot occur in real-time. Therefore, we incorporate the network-level data at an aggregate level, and within a probabilistic framework, in the trajectory planning process.
Not only has the CV technology enabled platooning, but it has provided the ground to allow for efficient, effective, and robust vehicle-to-platoon assignments as well as real-time management of platoons using both local and network-level information. This proposal introduces a holistic framework that allows for (i) finding stable platoon structures, and (ii) planning trajectories that would allow for forming such platoons, based on both fine-level local and coarse-level network-wide information.