TY - JOUR
T1 - Optimization of traffic flow at freeway sags by controlling the acceleration of vehicles equipped with in-car systems
AU - Goñi-Ros, Bernat
AU - Knoop, Victor L.
AU - Takahashi, Toshimichi
AU - Sakata, Ichiro
AU - van Arem, Bart
AU - Hoogendoorn, Serge P.
PY - 2016/10/1
Y1 - 2016/10/1
N2 - Sags are bottlenecks in freeway networks. According to previous research, the main cause is that most drivers do not accelerate enough at sags. Consequently, they keep longer headways than expected given their speed, which leads to congestion in high demand conditions. Nowadays, there is growing interest in the development of traffic control measures for sags based on the use of in-car systems. This paper aims to determine the optimal acceleration behavior of vehicles equipped with in-car systems at sags and the related effects on traffic flow, thereby laying the theoretical foundation for developing effective traffic management applications. We formulate an optimal control problem in which a centralized controller regulates the acceleration of some vehicles of a traffic stream moving along a single-lane freeway stretch with a sag. The control objective is to minimize total travel time. The problem is solved for scenarios with different numbers of controlled vehicles and positions in the stream, assuming low penetration rates. The results indicate that the optimal behavior involves performing a deceleration-acceleration-deceleration-acceleration (DADA) maneuver in the sag area. This maneuver induces the first vehicles located behind the controlled vehicle to accelerate fast along the vertical curve. As a result, traffic speed and flow at the end of the sag (bottleneck) increase for a time. The maneuver also triggers a stop-and-go wave that temporarily limits the inflow into the sag, slowing down the formation of congestion at the bottleneck. Moreover, in some cases controlled vehicles perform one or more deceleration-acceleration maneuvers upstream of the sag. This additional strategy is used to manage congestion so that inflow is regulated more effectively. Although we cannot guarantee global optimality, our findings reveal a potentially highly effective and innovative way to reduce congestion at sags, which could possibly be implemented using cooperative adaptive cruise control systems.
AB - Sags are bottlenecks in freeway networks. According to previous research, the main cause is that most drivers do not accelerate enough at sags. Consequently, they keep longer headways than expected given their speed, which leads to congestion in high demand conditions. Nowadays, there is growing interest in the development of traffic control measures for sags based on the use of in-car systems. This paper aims to determine the optimal acceleration behavior of vehicles equipped with in-car systems at sags and the related effects on traffic flow, thereby laying the theoretical foundation for developing effective traffic management applications. We formulate an optimal control problem in which a centralized controller regulates the acceleration of some vehicles of a traffic stream moving along a single-lane freeway stretch with a sag. The control objective is to minimize total travel time. The problem is solved for scenarios with different numbers of controlled vehicles and positions in the stream, assuming low penetration rates. The results indicate that the optimal behavior involves performing a deceleration-acceleration-deceleration-acceleration (DADA) maneuver in the sag area. This maneuver induces the first vehicles located behind the controlled vehicle to accelerate fast along the vertical curve. As a result, traffic speed and flow at the end of the sag (bottleneck) increase for a time. The maneuver also triggers a stop-and-go wave that temporarily limits the inflow into the sag, slowing down the formation of congestion at the bottleneck. Moreover, in some cases controlled vehicles perform one or more deceleration-acceleration maneuvers upstream of the sag. This additional strategy is used to manage congestion so that inflow is regulated more effectively. Although we cannot guarantee global optimality, our findings reveal a potentially highly effective and innovative way to reduce congestion at sags, which could possibly be implemented using cooperative adaptive cruise control systems.
KW - Car-following model
KW - In-car system
KW - Optimal control
KW - Sag vertical curve
KW - Stop-and-go wave
KW - Traffic management
UR - http://www.scopus.com/inward/record.url?scp=84978054069&partnerID=8YFLogxK
UR - http://resolver.tudelft.nl/uuid:61669680-bc59-4dd7-8b04-6730d6541468
U2 - 10.1016/j.trc.2016.06.022
DO - 10.1016/j.trc.2016.06.022
M3 - Article
SN - 0968-090X
VL - 71
SP - 1
EP - 18
JO - Transportation Research. Part C: Emerging Technologies
JF - Transportation Research. Part C: Emerging Technologies
ER -