Regional Multi-Agent Cooperative Reinforcement Learning for City-Level Traffic Grid Signal Control

This article studies the effective traffic signal control problem of multiple intersections in a city-level traffic system.A novel regional multi-agent cooperative reinforcement learning algorithm called RegionSTLight is proposed to improve the traffic efficiency.Firstly a regional multi-agent Q-learning framework is proposed,which can equivalently decompose the global Q value of the traffic system into the local values of several regions Based on the framework and the idea of human-machine cooperation,a dynamic zoning method is designed to divide the traffic network into several strong-coupled regions according to realtime traffic flow densities.In order to achieve better cooperation inside each region,a lightweight spatio-temporal fusion feature extraction network is designed.The experiments in synthetic real-world and city-level scenarios show that the proposed RegionS TLight converges more quickly,is more stable,and obtains better asymptotic performance compared to state-of-theart models.

An integrated and cooperative architecture for multi-intersection traffic signal control

Traffic signal control(TSC)systems are one essential component in intelligent transport systems.However,relevant studies are usually independent of the urban traffic simulation environment,collaborative TSC algorithms and traffic signal communication.In this paper,we propose(1)an integrated and cooperative Internet-of-Things architecture,namely General City Traffic Computing System(GCTCS),which simultaneously leverages an urban traffic simulation environment,TSC algorithms,and traffic signal communication;and(2)a general multi-agent reinforcement learning algorithm,namely General-MARL,considering cooperation and communication between traffic lights for multi-intersection TSC.In experiments,we demonstrate that the integrated and cooperative architecture of GCTCS is much closer to the real-life traffic environment.The General-MARL increases the average movement speed of vehicles in traffic by 23.2%while decreases the network latency by 11.7%.

Adaptive signal control and coordination for urban traffic control in a connected vehicle environment: A review

Existing signal control systems for urban traffic are usually based on traffic flow data from fixed location detectors.Because of rapid advances in emerging vehicular communication,connected vehicle(CV)-based signal control demonstrates significant improvements over existing conventional signal control systems.Though various CV-based signal control systems have been investigated in the past decades,these approaches still have many issues and drawbacks to overcome.We summarize typical components and structures of these existing CV-based urban traffic signal control systems and digest several important issues from the summarized vital concepts.Last,future research directions are discussed with some suggestions.We hope this survey can facilitate the connected and automated vehicle and transportation research community to efficiently approach next-generation urban traffic signal control methods and systems.

NSGA-Ⅱ based traffic signal control optimization algorithm for over-saturated intersection group

In order to improve the efficiency of traffic signal control for an over-saturated intersection group, a nondominated sorting genetic algorithm Ⅱ（NSGA-Ⅱ） based traffic signal control optimization algorithm is proposed. The throughput maximum and average queue ratio minimum for the critical route of the intersection group are selected as the optimization objectives of the traffic signal control for the over-saturated condition. The consequences of the efficiency between traffic signal timing plans generated by the proposed algorithm and a commonly utilized signal timing optimization software Synchro are compared in a VISSIM signal control application programming interfaces （SCAPI） simulation environment by using real filed observed traffic data. The simulation results indicate that the signal timing plan generated by the proposed algorithm is more efficient in managing oversaturated flows at intersection groups, and, thus, it has the capability of optimizing signal timing under the over-saturated conditions.

Fuzzy traffic signal control with DNA evolutionary algorithm

In order to optimize the signal control system, this paper proposes a method to design an optimized fuzzy logic controller （FLC） with the DNA evolutionary algorithm. Inspired by the DNA molecular operation characteristics, the DNA evolutionary algorithm modifies the corresponding genetic operators. Compared with the traditional genetic algorithm （GA）, the DNA evolutionary algorithm can overcome weak local search capability and premature convergence. The parameters of membership functions are optimized by adopting the quaternary encoding method and performing corresponding DNA genetic operators. The relevant optimized parameters are combined with the FLC for single intersection traffic signal control. Simulation experiments shows the better performance of the FLC with the DNA evolutionary algorithm optimization. The experimental results demonstrate the efficiency of the nrotmsed method.

Tri-level programming model for combined urban traffic signal control and traffic flow guidance

In order to balance the temporal-spatial distribution of urban traffic flow, a model is established for combined urban traffic signal control and traffic flow guidance. With consideration of the wide use of fixed signal control at intersections, traffic assignment under traffic flow guidance, and dynamic characteristics of urban traffic management, a tri-level programming model is presented. To reflect the impact of intersection delay on traffic assignment, the lower level model is set as a modified user equilibrium model. The middle level model, which contains several definitional constraints for different phase modes, is built for the traffic signal control optimization. To solve the problem of tide lane management, the upper level model is built up based on nonlinear 0-1 integer programming. A heuristic iterative optimization algorithm(HIOA) is set up to solve the tri-level programming model. The lower level model is solved by method of successive averages(MSA), the middle level model is solved by non-dominated sorting genetic algorithm II(NSGA II), and the upper level model is solved by genetic algorithm(GA). A case study is raised to show the efficiency and applicability of the proposed modelling and computing method.

Traffic Signals Control with Adaptive Fuzzy Controller in Urban Road Network

An adaptive fuzzy logic controller (AFC) is presented for the signal control of the urban traffic network. The AFC is composed of the signal control system-oriented control level and the signal controller-oriented fuzzy rules regulation level. The control level decides the signal timings in an intersection with a fuzzy logic controller. The regulation level optimizes the fuzzy rules by the Adaptive Rule Module in AFC according to both the system performance index in current control period and the traffic flows in the last one. Consequently the system performances are improved. A weight coefficient controller (WCC) is also developed to describe the interactions of traffic flow among the adjacent intersections. So the AFC combined with the WCC can be applied in a road network for signal timings. Simulations of the AFC on a real traffic scenario have been conducted. Simulation results indicate that the adaptive controller for traffic control shows better performance than the actuated one.

FL-FN-MOGA Based Traffic Signal Control

In this paper, a traffic signal control method based on fuzzy logic (FL), fuzzy-neuro (FN) and multi-objective genetic algorithms (MOGA) for an isolated four-approach intersection with through and left-turning movements is presented. This method has an adaptive signal timing ability, and can make adjustments to signal timing in response to observed changes.The 'urgency degree' term, which can describe the different user's demand for green time is used in decision-making by which strategy of signal timing can be determined. Using a fuzzy logic controller, we can determine whether to extend or terminate the current signal phase and select the sequences of phases. In this paper, a method based on fuzzy-neuro can be used to predict traffic parameters used in fuzzy logic controller. The feasibility of using a multi-objective genetic algorithm ( MOGA) to find a group of optimizing sets of parameters for fuzzy logic controller depending on different objects is also demonstrated. Simulation results show that the proposed methed is effecfive to adjust the signal timing in response to changing traffic conditions on a real-time basis, and the controller can produce lower vehicle delays and percentage of stopped vehicles than a traffic-actuated controller.

An ATMS data-driven method for signalized arterial coordination

In order to minimize the delays and stops caused by the early started coordinated green phase of the vehicle- actuated signal systems, a stochastic offsets calculation method based on the new types of advanced traffic management system （ATMS） data is proposed. As the mainline green starts randomly in vehicle-actuated signal systems, the random theory is applied to obtain the distribution of the unused green time at side streets based on the green gap-out mechanism. Then, the green start time of the mainline can be selected at the point with maximum probability to minimize the delays or stops caused by the randomly started mainline green. A case study in Maine, USA, whose traffic conditions are similar to those of the middle-size Chinese cities, proves that the proposed method can significantly reduce the travel time and delays.

Evolution of Road Traffic Congestion Control:A Survey from Perspective of Sensing,Communication,and Computation

Road traffic congestion can inevitably de-grade road infrastructure and decrease travel efficiency in urban traffic networks,which can be relieved by employing appropriate congestion control.Accord-ing to different developmental driving forces,in this paper,the evolution of road traffic congestion control is divided into two stages.The ever-growing num-ber of advanced sensing techniques can be seen as the key driving force of the first stage,called the sens-ing stage,in which congestion control strategies ex-perienced rapid growth owing to the accessibility of traffic data.At the second stage,i.e.,the communica-tion stage,communication and computation capabil-ity can be regarded as the identifying symbols for this stage,where the ability of collecting finer-grained in-sight into transportation and mobility reality improves dramatically with advances in vehicular networks,Big Data,and artificial intelligence.Specifically,as the pre-requisite for congestion control,in this paper,ex-isting congestion detection techniques are first elab-orated and classified.Then,a comprehensive survey of the recent advances for current congestion control strategies with a focus on traffic signal control,vehi-cle route guidance,and their combined techniques is provided.In this regard,the evolution of these strate-gies with continuous development of sensing,com-munication,and computation capability are also intro-duced.Finally,the paper concludes with several re-search challenges and trends to fully promote the in-tegration of advanced techniques for traffic congestion mitigation in transportation systems.

A multi process value-based reinforcement learning environment framework for adaptive traffic signal control

Realising adaptive traffic signal control(ATSC)through reinforcement learning(RL)is an important means to easetraffic congestion.This paper finds the computing power of the central processing unit(CPU)cannot fully usedwhen Simulation of Urban MObility(SUMO)is used as an environment simulator for RL.We propose a multi-process framework under value-basedRL.First,we propose a shared memory mechanism to improve exploration efficiency.Second,we use the weight sharing mechanism to solve the problem of asynchronous multi-process agents.We also explained the reason shared memory in ATSC does not lead to early local optima of the agent.Wehave verified in experiments the sampling efficiency of the 10-process method is 8.259 times that of the single process.The sampling efficiency of the 20-process method is 13.409 times that of the single process.Moreover,the agent can also converge to the optimal solution.

Adaptive green traffic signal controlling using vehicular communication

The importance of using adaptive traffic signal control for figuring out the unpredictable traffic congestion in today's metropolitan life cannot be overemphasized. The vehicular ad hoc network(VANET), as an integral component of intelligent transportation systems(ITSs), is a new potent technology that has recently gained the attention of academics to replace traditional instruments for providing information for adaptive traffic signal controlling systems(TSCSs). Meanwhile, the suggestions of VANET-based TSCS approaches have some weaknesses:(1) imperfect compatibility of signal timing algorithms with the obtained VANET-based data types, and(2) inefficient process of gathering and transmitting vehicle density information from the perspective of network quality of service(Qo S). This paper proposes an approach that reduces the aforementioned problems and improves the performance of TSCS by decreasing the vehicle waiting time, and subsequently their pollutant emissions at intersections. To achieve these goals, a combination of vehicle-to-vehicle(V2V) and vehicle-to-infrastructure(V2I) communications is used. The V2 V communication scheme incorporates the procedure of density calculation of vehicles in clusters, and V2 I communication is employed to transfer the computed density information and prioritized movements information to the road side traffic controller. The main traffic input for applying traffic assessment in this approach is the queue length of vehicle clusters at the intersections. The proposed approach is compared with one of the popular VANET-based related approaches called MC-DRIVE in addition to the traditional simple adaptive TSCS that uses the Webster method. The evaluation results show the superiority of the proposed approach based on both traffic and network Qo S criteria.

Urban Intersection Traffic Signal Control Based on Fuzzy Logic

This paper presents a fuzzy logic adaptive traffic signal control method for an isolated four-approach intersection with through and left-turning movements. In the proposed method, the fuzzy logic controller can make adjustments to signal timing in response to observed changes. The 'urgency degree' term that can describe different user's demands for a green light is used in the fuzzy logic decision-making. In addition, a three-level fuzzy controller model decides whether to extend or terminate the current signal phase and the sequence of phases. Simulation results show that the fuzzy controller can adjust its signal timing in response to changing traffic conditions on a real-time basis and that the proposed fuzzy logic controller leads to less vehicle delays and a lower percentage of stopped vehicles.

Image-based traffic signal control via world models

Traffic signal control is shifting from passive control to proactive control, which enables the controller to direct current traffic flow to reach its expected destinations. To this end, an effective prediction model is needed for signal controllers. What to predict, how to predict, and how to leverage the prediction for control policy optimization are critical problems for proactive traffic signal control. In this paper, we use an image that contains vehicle positions to describe intersection traffic states. Then, inspired by a model-based reinforcement learning method, DreamerV2,we introduce a novel learning-based traffic world model. The traffic world model that describes traffic dynamics in image form is used as an abstract alternative to the traffic environment to generate multi-step planning data for control policy optimization. In the execution phase, the optimized traffic controller directly outputs actions in real time based on abstract representations of traffic states, and the world model can also predict the impact of different control behaviors on future traffic conditions. Experimental results indicate that the traffic world model enables the optimized real-time control policy to outperform common baselines, and the model achieves accurate image-based prediction, showing promising applications in futuristic traffic signal control.

Traffic signal control in mixed traffic environment based on advance decision and reinforcement learning

Reinforcement learning-based traffic signal control systems (RLTSC) can enhance dynamic adaptability, save vehicle travelling timeand promote intersection capacity. However, the existing RLTSC methods do not consider the driver’s response time requirement, sothe systems often face efficiency limitations and implementation difficulties.We propose the advance decision-making reinforcementlearning traffic signal control (AD-RLTSC) algorithm to improve traffic efficiency while ensuring safety in mixed traffic environment.First, the relationship between the intersection perception range and the signal control period is established and the trust region state(TRS) is proposed. Then, the scalable state matrix is dynamically adjusted to decide the future signal light status. The decision will bedisplayed to the human-driven vehicles (HDVs) through the bi-countdown timer mechanism and sent to the nearby connected automatedvehicles (CAVs) using the wireless network rather than be executed immediately. HDVs and CAVs optimize the driving speedbased on the remaining green (or red) time. Besides, the Double Dueling Deep Q-learning Network algorithm is used for reinforcementlearning training;a standardized reward is proposed to enhance the performance of intersection control and prioritized experiencereplay is adopted to improve sample utilization. The experimental results on vehicle micro-behaviour and traffic macro-efficiencyshowed that the proposed AD-RLTSC algorithm can simultaneously improve both traffic efficiency and traffic flow stability.

Optimization of Adaptive Transit Signal Priority Using Parallel Genetic Algorithm

Optimization of adaptive traffic signal timing is one of the most complex problems in traffic control systems. This paper presents an adaptive transit signal priority （TSP） strategy that applies the parallel genetic algorithm （PGA） to optimize adaptive traffic signal control in the presence of TSP. The method can optimize the phase plan, cycle length, and green splits at isolated intersections with consideration for the performance of both the transit and the general vehicles. A VISSIM （VISual SIMulation） simulation testbed was developed to evaluate the performance of the proposed PGA-based adaptive traffic signal control with TSP. The simulation results show that the PGA-based optimizer for adaptive TSP outperformed the fully actuated NEMA control in all test cases. The results also show that the PGA-based optimizer can produce TSP timing plans that benefit the transit vehicles while minimizing the impact of TSP on the general vehicles.

Intersection signal control multi-objective optimization based on genetic algorithm

A signal control intersection increases not only vehicle delay, but also vehicle emissions and fuel consumption in that area. Because more and more fuel and air pollution problems arise recently, an intersection signal control optimization method which aims at reducing vehicle emissions, fuel consumption and vehicle delay is required heavily. This paper proposed a signal control multi-object optimization method to reduce vehicle emissions, fuel consumption and vehicle delay simultaneously at an intersection. The optimization method combined the Paramics microscopic traffic simulation software, Comprehensive Modal Emissions Model （CMEM）, and genetic algorithm. An intersection in Haizhu District, Guangzhou, was taken for a case study. The result of the case study shows the optimal timing scheme obtained from this method is better than the Webster timing scheme.

A fast network partition method for large-scale urban traffic networks

In order to control the large-scale urban traffic network through hierarchical or decentralized methods, it is necessary to exploit a network partition method, which should be both effective in extracting subnetworks and fast to compute. In this paper, a new approach to calculate the correlation degree, which determines the desire for interconnection between two adjacent intersections, is first proposed. It is used as a weight of a link in an urban traffic network, which considers both the physical characteristics and the dynamic traffic information of the link. Then, a fast network division approach by optimizing the modularity, which is a criterion to distinguish the quality of the partition results, is applied to identify the subnetworks for large-scale urban traffic networks. Finally, an application to a specified urban traffic network is investigated using the proposed algorithm. The results show that it is an effective and efficient method for partitioning urban traffic networks automatically in real world.