Adaptive Traffic Control System Research Articles

Application of LVQ Neural Network in Real–Time Adaptive Traffic Signal Control

Real–time road traffic data analysis is the cornerstone for the modern transport system. The real–time adaptive traffic signal control system is an essential part for the system. This analysis is to describe a traffic scene in a way similar to that of a human reporting the traffic status and the extraction of traffic parameters such as vehicle queue length, traffic volume, lane occupancy and speed measurement. This paper proposed the application of two–stage neural network in real–time adaptive traffic signal control system capable of analysing the traffic scene detected by video camera, processing the data, determining the traffic parameters and using the parameters to decide the control strategies. The two–stage neural network is used to process the traffic scene and decide the traffic control methods: optimum priority or optimum locality. Based on simulation in the traffic laboratory and field testing, the proposed control system is able to recognise the traffic pattern and enhance the traffic parameters thus easing traffic congestion more effectively than existing control systems. Key words: Urban traffic control system, pattern recognition, two-stage neural network, adaptive control system

Long-Term Benefits of Adaptive Traffic Control under Varying Traffic Flows during Weekday Peak Hours

When adaptive traffic control systems (ATCSs) are evaluated, traffic signal engineers and practitioners often collect data for a few weeks before and after installation. Benefits are then estimated on the basis of this limited data set. The evaluation of an ATCS with microsimulation requires considerable collection of field data. However, once an ATCS is installed, an abundance of data is collected and stored by the ATCS itself. These data (mostly traffic volumes) can be used to recreate field variability of traffic conditions in a model and perform long-term ATCS evaluation studies. This paper reports on the projected long-term benefits of deploying an ATCS. Field traffic data were statistically processed and modeled in microsimulation. The Sydney coordinated adaptive traffic system (SCATS) and two time-of-day (TOD) plans were exposed to variability of field traffic flows modeled in VISSIM. A simple calculation was then performed to extrapolate results to a period of 10 years. Findings showed that SCATS outperformed existing TOD signal-timing plans by about 20% and was better than the best TOD plan that could be theoretically developed on the basis of collection of long-term data. Results from the study revealed that short-term analyses often obscured the true benefits of deploying an ATCS. A computation of the monetary value of achieved benefits showed that limited operational benefits reported in this paper, when projected over the long term, would exceed overall installation costs for SCATS in Park City, Utah. These benefits are expected to increase further with inclusion of the analysis periods that were not part of this study.

Performance Evaluation of Intelligent Adaptive Traffic Control Systems: A Case Study

Mashhad, the second largest city in Iran, like many other big cities, is faced with increasing traffic congestion owing to rapidly increasing population and annual pilgrimage. In recent years, Mashhad traffic and transportation authorities have been challenged with how to manage the increasing congestion with limited budgets for major roadway construction projects. Mashhad has recognized the need to improve the existing system capacity to get the most out of their cur- rent transportation system infrastructures. Since most of the delay times occur at signalized intersections, using an intelligent control system with proper capabilities to overcome the growing traffic requirements is recommended. Following comprehensive studies carried out with the aim of developing the Mashhad traffic control center, the SCATS adaptive traffic control system was introduced as the selected intelligent control system for integrating signalized intersections. The first intersection was equipped with this system in 2005. This paper describes the results of a field evaluation in which fixed actuated-coordinated signal timings are compared with those dynamically computed by SCATS. The ef- fects of this system on optimizing fuel consumption as well as reducing air pollutants are fully discussed. It is found that SCATS consistently reduced travel times and the average delay per stopped or approaching vehicle. The positive impact of adaptive traffic control systems on fuel consumption and air pollution are also highlighted.

Evaluation of traffic signal control system using a system-wide performance measure under two-fluid model theory

This study compared the performance of an adaptive traffic control system and a semi-actuated coordinated control system using trip time and stop time relation under the two-fluid model theory. Using the test vehicle data collected for the evaluation of SCATS in Cobb County, Georgia, the parameters were estimated and interpreted. At the initial interpretation, potentially conflicting recommendations could be drawn from the individual route-period group level comparisons and the two-fluid model results. Therefore, a new interpretation combining the two approaches was presented. The new approach resulted in curves representing restricted ranges of the trip time versus stop time relationship. Using the curves and a two-fluid model assumption, the fraction of moving vehicles in the network can be obtained. Unlike the measures adopted in the previous traffic signal system evaluation studies, such as mean travel time and speed, this measure is defined by all time periods and routes data, allowing for a system-wide performance measure. Application of this measure showed that under SCATS approximately 10% more vehicles were moving on end-to-end type routes when the network was most congested and 15% to 20% less vehicles were moving on side-street to side-street type routes over different traffic concentration levels.

Arterial Performance Evaluation on an Adaptive Traffic Signal Control System

This paper documents the evaluation of Sydney Coordinated Adaptive Traffic System (SCATS) at a major signalized arterial in Las Vegas. The evaluation was based on various performance measures, including primarily travel times and stops at selected routes along the arterial. Extensive travel time runs were conducted during the weekday and weekend peak periods. Arterial performance was compared between SCATS and conventional time-of-day (TOD) coordination plans. Based on the data and analysis results, a general conclusion was reached that no significant improvement on arterial progression was achieved with SCATS under normal traffic conditions. However, this conclusion was made based on a major limitation of using video detection in the field, which may have significantly limited SCATS from achieving its best performance.

Methodology to Evaluate Active Transportation and Demand Management Strategies

This paper describes a methodology for predicting the impacts of active transportation and demand management (ATDM) measures on highway performance that can be applied using either microsimulation or Highway Capacity Manual (HCM) analysis techniques.Active transportation and demand management (ATDM) is a comprehensive approach to facility management and operation that seeks to increase facility productivity by proactively balancing supply and demand to avoid or delay facility breakdown. Examples of ATDM measures include: adaptive ramp metering, congestion pricing, speed harmonization, traveler information systems, and adaptive traffic signal control systems. Incident management and work zone management programs may employ one or more of these ATDM measures.The proposed ATDM analysis methodology is employed as a “wrapper” to a “core” analysis tool, such as the HCM or microsimulation. The core analysis tool is applied several times to a set of demand/capacity scenarios. The demand/capacity scenarios are developed based on observations of annual fluctuations in weekday peak period demands and capacities for the facility. The scenario results are then combined according to their probability of occurrence in order to predict the full-year peak period performance of the street, highway, or freeway facility.The application of the methodology is illustrated on a real-world freeway facility. The Highway Capacity Manual FREEVAL procedure, the selected “core” analysis tool, was adapted to test ramp metering, HOV (High Occupancy Vehicle) lanes, and HOT (High Occupancy Toll) lanes within the ATDM evaluation methodology. FREEVAL was applied to each of several demand/capacity scenarios and the results combined into an estimate of average performance improvements for each ATDM measure. The predicted facility performance improvements were compared to a before/after field studies reported in the literature. © 2011 Published by Elsevier Ltd.

Comparative Evaluation of Adaptive Traffic Control System Assessments Through Field and Microsimulation

Usually, benefits from deploying an adaptive traffic control system are measured by comparing data collected in the field before the system is installed with those collected after the system is fully operational. Such an approach is costly because it requires installation of a system before potential benefits can be observed. Nowadays to investigate effectiveness of many alternatives that were traditionally investigated in field, microsimulation is used. However, how much can one rely on results from microsimulation? There has not been a single study in which benefits of adaptive traffic control systems measured in the field are compared with those obtained through microsimulation. This study presents such a research effort. The authors compared performance measures from a field evaluation of Sydney Coordinated Adaptive Traffic System in Park City, Utah, to their counterparts from microsimulation. They collected a significant amount of data to assess system performance in the field and to help build a microsimulation model. A 14-intersection model of the Park City network was developed, calibrated, and validated on the basis of multiple data sources. The results show that a validated microsimulation model can accurately reflect field conditions although such an effort can be very challenging.

Towards multi-agent reinforcement learning for integrated network of optimal traffic controllers (MARLIN-OTC)

Traffic congestion can be alleviated by infrastructure expansions; however, improving the existing infrastructure using traffic control is more plausible due to the obvious financial resources and physical space constraints. The most promising control tools include ramp metering, variable message signs, and signalized intersections. Synergizing the aforementioned strategies in one platform is an ultimate and challenging goal to alleviate traffic gridlock and optimally utilize the existing system capacity; this is referred to as <i>Integrated Traffic Control (ITC). Reinforcement Learning (RL)</i> techniques have the potential to tackle the optimal traffic control problem. <i>Game Theory (GT)</i> fits well in modelling the distributed control systems as multiplayer games. <i>Multi-Agent Reinforcement Learning (MARL)</i> achieves the potential synergy of <i>RL</i> and <i>GT</i> concepts, providing a promising tool for optimal distributed traffic control. The objective of this paper is to clarify the opportunities of game theory concepts and <i>MARL</i> approaches in creating an adaptive optimal traffic control system that is decentralized but yet integrated through agents' interactions. In this paper, we comparatively review and evaluate the relevant existing approaches. We then envision and introduce a novel framework that combines <i>GT</i> concepts and <i>MARL</i> to achieve a <i>Multi-Agent Reinforcement Learning for Integrated Network of Optimal Traffic Controllers (MARLIN-OTC).</i>

Managing Congestion in Large Brazilian Urban Area through Logical Interface between SCOOT and GIS Platform

Effective management of traffic congestion is based on a continuous, systematic understanding of travel behavior in the urban network. A continuous effort of data collection, treatment, and analysis is demanded and requires time, financial resources, and skilled staff. An efficient alternative for urban traffic management consists of implementing traffic management centers (TMCs), which collect, model, and store data relating to traffic conditions. However, the operational systems usually implemented in those centers do not offer effective mechanisms for consulting their databases because they are designed with an emphasis on traffic signal control. This work presents the development of a logical interface between the dynamic database of the TMC of Fortaleza, Brazil (CTAFOR), and the TransCAD geographic information system (GIS) platform. This interface combines an open modular architecture and a spatiotemporal model to import dynamic traffic data into the static GIS platform. The interface also provides flexible, efficient access to the data collected in real time by CTAFOR's adaptive traffic control system, SCOOT, and gives the data a spatial attribute in TransCAD software. With the interface available, the second main objective was to implement performance measures to evaluate the efficiency and effectiveness of the operational actions taken in the area controlled by CTAFOR and assess how well the objectives set to the TMC are being achieved. After the development of the TRANSCOOT interface, the ITS data collected and simulated at CTAFOR have been used to support real-time operations and to allow tactical, strategic decision analyses, based on performance measures aggregated by their spatial and temporal attributes.

Comparison of Before–After versus Off–On Adaptive Traffic Control Evaluations

An adaptive traffic control system, the Sydney Coordinated Adaptive Traffic System (SCATS), was installed in Park City, Utah, to improve traffic performance at its network of signalized intersections. A field evaluation of the previous time-of-day actuated–coordinated signal timings was conducted before SCATS installation to compare the two systems. However, the post-SCATS field evaluation could not occur until two additional signals were installed and several other changes were made to the network. Two years after the original pre-SCATS field evaluation, the network was reevaluated with an off–on technique analogous to a before–after study. The signal timings and parameters in the off condition forced SCATS to use time-of-day actuated–coordinated control, similar to the before study but with timings readjusted for the additional signals and changed traffic conditions. The performance gains with SCATS on were measurably greater than those with SCATS off for travel time and number of stops and greater overall for stopped delay. These data provided the transportation agency with results anticipated by its traffic engineers. However, the original field evaluation data from 2 years earlier provided a less distinct conclusion. A methodology is presented to determine the relevance of an off–on study in place of a before–after study. The results show that “before” and “off ” data sets behave consistently 62.5% of the time. This value provides a basis of support for using off data, which better represent before signal timings on an after network. It also quantifies sensitivity to changes in the network, which are substantial in this case.

Neural-network-based cycle length design for real-time traffic control

Adaptive traffic control systems (ATCS) are designed to calculate traffic signal timings in real time to accommodate current traffic demand changes. A conventional off-line computer-based design procedure that uses iterative evaluations to select alternatives may not be appropriate for ATCS due to its unstable searching time. Search-free analytical procedures that directly find solutions have been noted for ATCS for this reason. This paper demonstrates (i) the shortcomings of an analytical cycle-length design model, specifically COSMOS, in its ability to generate satisfactory solutions at various saturation levels and (ii) an artificial neural network (ANN) based model that can overcome these shortcomings. The ANN-based model consistently yielded cycle lengths that ensure a proper operational target volume to capacity (v/c) ratio, whereas the use of the analytical model resulted in unstable target v/c ratios that might promote congestion.

TRAFFIC ADAPTIVE CONTROL OF A SINGLE INTERSECTION: A TAXONOMY OF APPROACHES

The design of a traffic adaptive control system is as well a science as an art. Along the way compromises have to be made in order to end up with a workable system that is not only able to come up with good signal timings, but is also able to deliver them on time. In this paper we propose a taxonomy of the various traffic adaptive control algorithms based both on their underlying principles and the compromises that were made to come up with a workable, albeit less optimal system.

An Enhanced 0–1 Mixed-Integer LP Formulation for Traffic Signal Control

An enhanced 0-1 mixed-integer linear programming formulation based on the cell-transmission model is proposed for the traffic signal optimization problem. This formulation has several features that are currently unavailable in other existing models developed with a similar approach, including the components for handling the number of stops, fixed or dynamic cycle length and splits, and lost time. The problem of unintended vehicle holding, which is common in analytical models, is explicitly treated. The formulation can be utilized in developing strategies for adaptive traffic-control systems. It can also be used as a benchmark for examining the convergence behavior of heuristic algorithms based on the genetic algorithm, fuzzy logic, neural networks, or other approaches that are commonly used in this field. The discussion of extending the proposed model to capture traffic signal preemption in the presence of emergency vehicles is given. In terms of computational efficiency, the proposed formulation has the least number of binary integers as compared with other existing formulations that were developed with the same approach.

Single-Intersection Evaluation of Real-Time Adaptive Traffic Signal Control Algorithms

Adaptive traffic signal control systems present the potential to significantly reduce delay, travel time, stops, and fuel consumption relative to current arterial traffic management practices. However, this technology—now referred to as adaptive-control software (ACS) by FHWA—has not been widely deployed in the United States, in part because traffic engineers are unconvinced of the incremental benefits of ACS over those of well-maintained time-of-day plans. This study identifies three specific deficiencies in the current generation of adaptive-control algorithms. Primarily, the inability to optimize timing in general multiring configurations prevents current adaptive-control algorithms from consistently out-performing properly calibrated dual-ring vehicle-actuated controllers. Limitations in real-time computational capabilities are identified as the main roadblock to overcoming this and other deficiencies. This limitation motivates a novel simulation study focused on real-time computational capabilities of adaptive-control algorithms used in OPAC, PRODYN, ALLONS-D, and COP, the intersection control algorithm of RHODES. The goal is to determine which algorithmic techniques are best suited for extension or reformulation to overcome the identified deficiencies. Benchmarks for vehicle delay performance include Webster's optimized fixed-time control, vehicle-actuated control, and an adaptive fixed-time policy derived from SCOOT. In conclusion, certain aspects of algorithmic techniques from OPAC, PRODYN, and ALLONS-D are unsuitable for much more computationally demanding tasks, including future extension to full optimization of a multiring controller. Also, specific features of the COP-97 and ALLONS-D algorithms present significant advantages in computational complexity and reduced traffic delay, respectively.

Cooperative, hybrid agent architecture for real-time traffic signal control

This paper presents a new hybrid, synergistic approach in applying computational intelligence concepts to implement a cooperative, hierarchical, multiagent system for real-time traffic signal control of a complex traffic network. The large-scale traffic signal control problem is divided into various subproblems, and each subproblem is handled by an intelligent agent with a fuzzy neural decision-making module. The decisions made by lower-level agents are mediated by their respective higher-level agents. Through adopting a cooperative distributed problem solving approach, coordinated control by the agents is achieved. In order for the multiagent architecture to adapt itself continuously to the dynamically changing problem domain, a multistage online learning process for each agent is implemented involving reinforcement learning, learning rate and weight adjustment as well as dynamic update of fuzzy relations using an evolutionary algorithm. The test bed used for this research is a section of the Central Business District of Singapore. The performance of the proposed multiagent architecture is evaluated against the set of signal plans used by the current real-time adaptive traffic control system. The multiagent architecture produces significant improvements in the conditions of the traffic network, reducing the total mean delay by 40% and total vehicle stoppage time by 50%.

Adaptive traffic control system: Control strategy, prediction, resolution, and accuracy

AbstractRecent advances in traffic control methods have led to flexible control strategies for use in an adaptive traffic control system (ATCS). ATCS aims at controlling the imminent traffic, which is yet to arrive and hence not known perfectly. Therefore, volume prediction is an essential part. Associated with the prediction are two aspects: resolution and accuracy. Recent studies indicate a tradeoff between prediction resolution and accuracy: finer resolutions, larger errors. It is imperative to study the relationship and tradeoff between the control strategy, prediction resolution, and its associated error, which are crucial to the development of ATCS. This study investigates this relationship through an extensive simulation of scenarios in Hong Kong with a recently developed dynamic traffic control model, DISCO. Based on the Hong Kong scenarios conducted with DISCO, the major findings include: (i) the importance of resolution outweighs that of error; (ii) dynamic timing plans generally outperform time‐invariant timing plans; (iii) up to a certain extent, overestimated predictions lead to better results than underestimated predictions.

Development of degree of saturation estimation models for adaptive signal systems

Throughout the world, several areawide traffic control systems have been developed and implemented. The degree of saturation plays a crucial role for the optimization of signal control variables. This paper investigates an alternate way to estimate the degree of saturation. From a correlation analysis, a strong positive linear relationship was found between percent occupancy time and degree of saturation. Based on this result, a linear regression model was constructed. In addition, a new model was also suggested for the use of traffic control systems. This paper also presents comparative results on degree of saturation models adopted in several adaptive traffic control systems. Each model was tested and evaluated using field data in terms of estimation errors. From the results, it was observed that SCOOT and SCATS models produced better estimates than other models under the light flow conditions. Both the new and regression models underestimated slightly for all periods, whereas the Barcelona model overestimated consistently. Under the heavy flow conditions, however, the new model showed improved performance over the other models.

Optimized Policies for Adaptive Control Strategy in Real-Time Traffic Adaptive Control Systems: Implementation and Field Testing

The real-time traffic adaptive control system (RT-TRACS) represents a new, state-of-the-art system in advanced traffic signal control. It was developed cooperatively by a team of U.S. academic, private, and public researchers under the guidance of FHWA. The system provides a framework to run multiple traffic control algorithms, existing ones as well as new adaptive algorithms, as they become available. The optimized policies for adaptive control (OPAC) strategy, which provides a dual capability of individual intersection control as well as coordinated control of intersections in a network, was the first adaptive algorithm implemented within the RT-TRACS framework. The operational features of the OPAC prototype version that was developed for RT-TRACS are presented, and its implementation in the Reston Parkway field research test bed in Northern Virginia is described. The implementation provided valuable insights into the performance of coordinated OPAC under various traffic conditions and site geometry. Observations indicate that the strategy was instrumental in reducing delays and stops, compared with a well-tuned fixed time system that was in place, while maintaining progression along the arterial. Valuable lessons were learned that should lead to improvements in future implementations of adaptive algorithms.

Remote Simulation to Evaluate Real-Time Traffic Control Strategies

Simulation is a valuable tool for evaluating the effects of various changes in a transportation system. This is especially true in the case of real-time traffic-adaptive control systems, which must undergo extensive testing in a laboratory setting before being implemented in a field environment. Various types of simulation environments are available, from software-only to hardware-in-the-loop simulations, each of which has a role to play in the implementation of a traffic control system. The RHODES (real-time hierarchical optimized distributed effective system) real-time traffic-adaptive control system was followed as it progressed from a laboratory project toward actual field implementation. The traditional software-only simulation environment and extensions to a hardware-in-the-loop simulation are presented in describing the migration of RHODES onto the traffic controller hardware itself. In addition, a new enhancement to the standard software-only simulation that allows remote access is described. The enhancement removes the requirement that both the simulation and the traffic control scheme reside locally. This architecture is capable of supporting any traffic simulation package that satisfies specific input-output data requirements. This remote simulation environment was tested with several different types of networks and was found to perform in the same manner as its local counterpart. Remote simulation has all of the advantages of its local counterpart, such as control and flexibility, with the added benefit of distribution. This remote environment could be used in many different ways and by different groups or individuals, including state or local transportation agencies interested in performing their own evaluations of alternative traffic control systems.

Impact of Adaptive Signal Control on Major and Minor Approach Delay

One of the primary difficulties in completing evaluations of real-time adaptive traffic signal systems has been the lack of effective modeling tools to perform controlled and repeatable analytical experiments. Recently, a method to analyze the delay characteristics of an existing real-time adaptive traffic signal control system has been developed. The procedure has been used to evaluate the performance of the Sydney Coordinated Adaptive Traffic System (SCATS) in South Lyon, Michigan. The objective of this study was to more closely examine the relationship of major and minor street delay under the SCATS and fixed-time forms of signal control. The effect on left turn delay at critical locations was also compared to the delay change in the through traffic. The comparisons showed that SCATS control resulted in significant delay reductions to left turn traffic compared to the through movements. The improvement was especially pronounced at the major street approaches. To further address the delay effect of SCATS control in South Lyon, the relationships between volume, delay, and green time of the various traffic moments were also compared. These comparisons showed that SCATS tended to allocate more green time to left turn traffic when compared to the fixed-time system.