A group-based adaptive traffic-control method for isolated signalized junctions is developed that includes a hierarchical structure comprising tactical and local levels of signal timing optimization. The control method optimizes the signal timings in adaptive traffic-control systems, and takes full advantage of flexible new technologies to incorporate the most up-to-date traffic information, as collected in real time. A multi-resolution strategy is proposed for updating the elements of the signal plans cycle-by-cycle and adjusting the current green signal timing second-by-second. The definitions, combinations, and sequencing of the cycle structure stages are generated automatically using a procedure for optimizing the signal-timing plans in response to online data from traffic detectors. There is a high degree of flexibility in the tactical identification of the optimal signal plan in response to the real-time predicted traffic information, the objective function of the polygonal delay formula, and the direct differential equations for the adaptive group-based variables. The reactive local signal-control policy, which is formed based on the max-pressure strategy, is developed to locally adjust the current green signal time and to accommodate delicate demand fluctuations second-by-second at the fine-resolution level. The results of the computer simulations presented in this study show that the integrated group-based adaptive traffic-signal-control logic outperforms the other methods over a wide range of traffic conditions, from free-flowing traffic to extreme congestion. Moreover, the proposed models perform much better than the existing fixed-signal plan and the actuated signal-control in asymmetric traffic conditions.

The occurrence of Connected and Automated Vehicles (CAV) technologies brings an opportunity to develop an online data-driven signal coordination method which cooperatively optimizes the traffic performance of a group intersections. This research proposes a game theoretic approach to epsilon-equilibrium (or near-Nash equilibrium) to achieve online traffic signal coordination. Each intersection of the traffic network is just like a player in a game. Different intersections pursue their own benefit maximization by changing their offsets. This paper examines the effect, applicability, and efficiency of the game theoretic approach in signal coordination. The game theoretical approach is proven to outperform the system optimum on vehicle delay at intersection level regarding delay equity. The variances of vehicle delay among different intersections are significantly decreased by the proposed game theoretic algorithm. Thus no intersection needs to sacrifice its own delay performance to achieve system optimum, and traffic delay has been widely distributed among intersections. In addition, this research also compares the network delay performances between CAVs and Human-driven Vehicle (HDV). A simulation platform is built to evaluate the proposed algorithms and models. The results also show that the CAVs can generate much less delay compared to HDVs.

We propose a novel traffic control algorithm that can be used in oversaturated road networks in large cities. For sophisticated traffic control, it is important to consider the dynamic behavior of vehicle queues. Such queues obstruct traffic flow in various directions, making the traffic situation chaotic. This results in severe traffic congestion, or so-called gridlock. In this study, we focus on closed-loop structures in road networks and vehicle queue advancement in these loops. We formalize the occurrence of gridlock, analyzing this condition, and we propose a traffic control algorithm called Z-control, which minimizes the total time spent based on a model predictive scheme. Moreover, we evaluated some algorithms for dynamic traffic control, including our proposed Z-control and some benchmarks. To do so, we conducted numerical experiments, considering two extreme scenarios pertaining to the behavior of drivers. In a “non-reactive” scenario, where drivers do not react to traffic conditions in the network, we showed that the proposed Z-control outperformed other benchmark algorithms. However, in a “reactive” scenario, where drivers are assumed to have perfect information about the traffic conditions in the network and react to it, no algorithm performed better than the situation without any control. Our research shows that to treat oversaturated traffic in cities suitably, it is important to control, or at least consider, the route-choice behavior of drivers, which may be achieved with automated vehicles.

This paper presents the results of a new deep learning model for traffic signal control. In this model, a novel state space approach is proposed to capture the main attributes of the control environment and the underlying temporal traffic movement patterns, including time of day, day of the week, signal status, and queue lengths. The performance of the model was examined over nine weeks of simulated data on a single intersection and compared to a semi-actuated and fixed time traffic controller. The simulation analysis shows an average delay reductions of 32% when compared to actuated control and 37% when compared to fixed time control. The results highlight the potential for deep reinforcement learning as a signal control optimization method.

In this paper, a decentralized traffic signal control strategy named max pressure control is reviewed and examined. This control strategy aims at optimizing overall network throughputs, but applies a distributed approach that only requires local information to generate timing plans for each intersection. A Vissim simulation study has been conducted to compare existing max pressure schemes. The results show that a recently proposed cyclic-based approach performs more poorly than the original non-cyclic approach. Further, to address frequent changes of phases and queue spillover/blockage, two issues that existing schemes suffer, two modifications are suggested. The simulation reveals that network performance can be improved after modifications.

This paper presents a novel de-centralized traffic signal controller, achieved using a Nash bargaining game-theoretic framework, that operates a flexible phasing sequence to adapt to dynamic changes in traffic demand. The Nash bargaining algorithm optimizes each signalized intersection by modeling each phase as a player in a game in which the players cooperate to reach a mutual agreement. The proposed algorithm was implemented and evaluated using the INTEGRATION microscopic traffic assignment and simulation software.
The proposed control approach is compared to the operation of an optimum fixed-time coordinated plan, an actuated controller, a centralized adaptive phase split controller, a decentralized phase split and cycle length controller, and a fully coordinated adaptive phase split, cycle length, and offset optimization controller to evaluate the performance of the proposed decentralized controller. The proposed controller was tested on an isolated intersection, showing a 77% reduction in queue length, 17% reduction in vehicle emission levels, 65% reduction in average travel time, and 64% reduction in the total delay. In the case of the arterial network the proposed algorithm produced reductions in total delay ranging from 36% to 67%, in the average travel time ranging from 7% to 21%, and a reduction in vehicle emissions ranging from 6% to 13%, which were demonstrated to be statistically significant. These results demonstrate that major benefits that can be achieved from the proposed de-centralized controller.

In some form or another, model predictive traffic control has been proposed for some decades. Because it predicts future traffic conditions, rather than reacting to past conditions in hindsight, it has advantages over adaptive traffic control techniques. However, it has not been widely adopted for a couple of reasons. Firstly, there are doubts about its suitability when fine-grained traffic data is unavailable. This objection can be addressed with the promise of vehicle-to-infrastructure (V2I) communication. Secondly, previous implementations have found that the control algorithm’s computational complexity is exponential or worse. This paper addresses this objection by introducing the A* algorithm to significantly decrease computation time. Other methods for reducing computation time include setting a suitable prediction horizon, clustering vehicles into platoons and implementing an incremental version of the control algorithm. With these methods combined, it is shown through simulation, that a real-time model predictive control algorithm is practical because computation time is manageable without having an adverse effect on total delay.

This paper proposes a double pressure (DP) signal control strategy that takes into account finite link storage capacity. Two types of pressure measures, including an endogenous and an exogenous pressure, were defined. The endogenous pressure is associated with the estimated exit-flow during a signal phase in the next time period, while the exogenous pressure is a virtual pressure associated with the downstream links. The properties of the proposed signal control strategy, including waiting time minimization and queue length stabilization were discussed. The performance of the proposed signal control strategy was evaluated with two tested networks via simulation, including a network with an isolated signalized intersection and a network with multiple signalized intersections. We also compared the performance of the proposed DP control strategy with that of a backpressure algorithm based (BP) control strategy proposed by previous studies. The test results suggest that the proposed DP control strategy outperforms the BP control strategy in terms of lower network travel time, lager network throughputs, and better queue stability performance in most traffic-volume conditions.

This paper compares the performance of a traffic responsive intersection controller which combines vehicle Global Positioning System (GPS) data and inductive loop information, to fixed-time, inductive loop, and GPS based controllers.
The INRIX Global Traffic Scorecard reports that road users spent up to 42% of their time in congested traffic in 2016.
Inefficient signal timing choices by isolated intersection controllers contribute to traffic delays, causing severe negative impacts on the economy and environment.
Signal timings can be improved using vehicles' GPS information combined with vehicle flow information from inductive loops to overcome the control action deficit at isolated intersections.
This proposed new signal control algorithm is beneficial for traffic engineers and governmental agencies, as optimised traffic flow can reduce fuel consumption and emissions.

The proposed traffic responsive Hybrid Vehicle Actuation (HVA) algorithm uses position and heading data from vehicle status broadcasts, and inferred velocity information to determine vehicle queue lengths and detect vehicles passing through the intersection to actuate intersection signal timings. When vehicle broadcast data are unavailable, HVA uses inductive loop data.
Microscopic simulations comparing HVA to fixed-time control, inductive Loop Based Vehicle Actuation (Loop-VA) and GPS Based Vehicle Actuation (GPS-VA) on four urban road networks were carried out to see how the proposed HVA algorithm performs compared to existing control strategies.
The results show that HVA is an effective alternative to traditional intersection control strategies, offering delay reductions of up to 32% over Loop-VA, for networks with 0-100% connected vehicle presence.

This research proposes an adaptive actor-critic reinforcement agent, trained using the asynchronous advantage actor-critic algorithm, as an improved traffic signal controller. Using an aggregate statistic state representation (i.e., vehicle queue and density), the reinforcement learning traffic signal controller develops the optimal policy compared to traditional, Webster’s and loop-detector actuated, traffic signal control methods in two out of three stochastic environments. The agent is able to learn a traffic signal control policy that exhibits generalization at traffic demands below and at capacity but not above capacity.

In this paper, we study the problem of devising a signal control policy for road transportation networks that is fair and provides high throughput. In particular, we propose and study a novel Queue-Delay Backpressure algorithm with variable cycle length, that takes into account both the queue lengths and the head-of-line delay at a junction. Using a variety of simulations, we show that the proposed algorithm achieves a middle ground between maximizing throughput and minimizing the maximum delays incurred. We then optimize cycle lengths to achieve minimum weighted sum of delays. Finally, we also study the effect of explicitly considering start-up losses in headways while optimizing the cycle lengths, and conclude that it is beneficial to consider these effects while designing signal control policies.

Many efforts have been carried out to optimize the traffic signal settings in 
cities. Nevertheless, state-of-the-art and -practice strategies cannot deal 
efficiently with oversaturated conditions (i.e. queue spillbacks and partial 
gridlocks), as they are either based on application-specific heuristics or 
they fail to replicate accurately the propagation of congestion. An 
alternative approach for real-time network-wide control is the perimeter flow 
control (or gating). This can be viewed as an upper-level control layer, and 
be combined with other strategies (e.g. local or coordinated regulators) in a 
hierarchical control framework. In the current work, a recently developed 
perimeter control regulator is utilized for the upper-level layer. Another
     lower-level control layer utilizes the max-pressure regulator, which 
constitutes a local feedback control law, applied in coupled intersections, 
in a distributed systems-of-systems (SoS) concept. Different approaches are 
discussed about the design of the hierarchical structure of SoS and a traffic 
microsimulation tool is used to assess the impact of each approach to the 
 overall traffic conditions. Preliminary results show that integrating a 
    network-level approach within a local adaptive framework can significantly 
  improve the system performance when spillback phenomena occur (a common 
feature of city centres with short links).

The performance of arterial corridors is significantly determined by the signal control systems at intersections. Currently, most research effort has been spent to investigate adaptive signal control systems with advanced technologies, such as connected vehicles. However, these systems cannot be easily realized due to the limitation of the technologies. This paper aims to develop an innovative platoon-based signal control using the existing platform of actuated signal controller to improve the performance of intersection. The system utilizes existing road facilities, such as loop detectors and cameras, to monitor the arrivals of vehicle platoons from different approaches of intersections, and to search for the optimal signal phasing and timing plan. The specialty of this system is that it applies the conventional signal control equipment making the application in real world less expensive and
sustainable. A VISSIM simulation is applied to realize the application and evaluate its effectiveness on reducing traffic congestion at intersection. The system is further applied at one intersection in Houston, Texas to illustrate its real-world benefits. The field experiment indicates that the system reduce
vehicle delay about 11% and the average queue length up to 60%.

ABSTRACT

Adaptive signal control systems are widely used in big cities around the world. Most of the current systems decide based on a set of logics to extend or cut the current signal stage (phase) in order to maximize the road network use. With the advancements of computers and optimization methods, new systems have been designed. These systems optimize the timings by applying a heuristic optimization method to a target function that contains different traffic measures for each mode of transport.

In this paper, a comparison is made in VISSIM traffic micro-simulator between the widely used system SCATS (Version 6.5) as a logic-based adaptive signal control and new developed BALANCE/EPICS as an optimization-based one. A part of the CBD of Tehran with three signalized intersections was modeled in VISSIM from 06:00 to 10:00 AM with around 44,000 veh-trips. Additionally 322 BRT vehicles cross the West intersection. The area was already equipped with the SCATS. To make the comparison, the signal timings of the SCATS in VISSIM were replaced with BALANCE/EPICS Signal Controllers.

The results show significant improvement of the performance after applying the optimization-based system. The difference is more considerable during the peak time. Delay, queue length and fuel consumption are decreased by 37%, 45% and 24% respectively. The delay of the BRT vehicles is decreased by 48%.

It should be clarified that there are some critiques to the methodology that are described in the paper.

Keywords: Adaptive signal control system, Heuristic optimization, SCATS, BALANCE, EPICS, Micro-simulation, VISSIM