Intersection Management via the Opportunistic Organization of Platoons by Route

TitleIntersection Management via the Opportunistic Organization of Platoons by Route
Publication TypeConference Paper
Year of Publication2015
AuthorsHeras, A, Claytor, L, Volos, H, Asadi, H, Sprinkle, J, Bose, T
Conference NameWinnComm 2016
AbstractIntersections behave as a pathway for traffic redirection and play an important role in the general movement of traffic, because of this, traffic congestion is not desirable and methods to remove congestion or its effects have been implemented. Generally, congestion occurs when a group of vehicles that is large relative to available road space attempts to move through an area that cannot redirect the group effectively. The are of this paper’s focus is the intersection. Some methods currently implemented to control congestion are roadway expansion, the use of adaptive traffic lights, and the use of public transportation. Good short term solutions, they only postpone the occurrence of congestion and are also costly. One way to deal with congestion is have vehicles cooperate with each other and their intersections through means of wireless communication. The vehicles can then be organized by a management system in such a way that vehicle throughput at an intersection can be maintained even at high levels of traffic congestion. A requirement of the system proposed in this paper is that all vehicles behave autonomously in order to obtain easily predictable behavior. The following describes the results of such a simulation of such a system. The data gathered over four simulations show that a higher platoon depth causes a general decrease in delay. The results between signalized and un-signalized simulations were vastly different with the un-signalized having significantly better results. The delay at a platoon depth of 9, for the unsignalized intersection, was -1.8, this brings the ideal delay to 98.2 seconds down from 100 seconds. This was due to the way intersection merging and gap threading is performed and is valid for several reasons. Because this method of intersection management desires to decrease the delay in traffic as much as possible, it would not make sense to slow down a platoon in order to perform a maneuver. The signalized method performed significantly worse than the un-signalized due to how the signalized method was implemented. Only one traffic light was green at a time, hence, whenever a platoon was stopped at intersection, that platoon would be guaranteed a delay of at most 56 seconds, which is how long a red light lasts for one signal. Though inefficient signal timing was used, we hypothesize that the behavior of a right of way system, such as a traffic light, is fundamentally flawed. By delegating the right of way to only a few lanes at a time, it is guaranteed that some vehicles will be stopped and will therefore experience delay. In order to show that the capacity of signalized intersection management through the use of traffic lights falls short when compared to un-signalized autonomous intersection management, better traffic light control methods will need to be tested in the future.