Traffic-actuated Control Research Articles

Machine Learning Model Application and Comparison in Actuated Traffic Signal Forecasting.

Traffic signal forecasting plays a significant role in intelligent traffic systems since it can predict upcoming traffic signal without using traditional radio-based direct communication with infrastructures, which causes high risk in the communication security. Previously, mathematical and statistical approach has been adopted to predict fixed time traffic signals, but it is no longer suitable for modern traffic-actuated control systems, where signals are dependent on the dynamic requests from traffic flows. And as a large amount of data is available, machine learning methods attract more and more attention. This paper views signal forecasting as a time-series problem. Firstly, a large amount of real data is collected by detectors implemented at an intersection in Hanover via IoT communication among infrastructures. Then, Baseline Model, Dense Model, Linear Model, Convolutional Neural Network, and Long Short-Term Memory (LSTM) machine learning models are trained by one-day data and the results are compared. At last, LSTM is selected for a further training with one-month data producing a test accuracy over 95%, and the median of deviation is only 2 s. Moreover, LSTM is further evaluated as a binary classifier, generating a classification accuracy over 92% and AUC close to 1.

A Robust Adaptive Traffic Signal Control Algorithm Using Q-Learning under Mixed Traffic Flow

The operational and safety performance of intersections is the key to ensuring the efficient operation of urban traffic. With the development of automated driving technologies, the ability of adaptive traffic signal control has been improved according to data detected by connected and automated vehicles (CAVs). In this paper, an adaptive traffic signal control was proposed to optimize the operational and safety performance of the intersection. The proposed algorithm based on Q-learning considers the data detected by loop detectors and CAVs. Furthermore, a comprehensive analysis was conducted to verify the performance of the proposed algorithm. The results show that the average delay and conflict rate have been significantly optimized compared with fixed timing and traffic actuated control. In addition, the performance of the proposed algorithm is good in the test of the irregular intersection. The algorithm provides a new idea for the intelligent management of isolated intersections under the condition of mixed traffic flow. It provides a research basis for the collaborative control of multiple intersections.

Maximum Green Time Settings for Traffic-Actuated Signal Control at Isolated Intersections Using Fuzzy Logic

Traffic Actuated Control (TAC) is widely used for signal control in isolated intersections. TAC is a responsible method for online traffic data. In TAC, green durations are dynamically determined based on online traffic data. These determinations are subject to predefined static parameters which do not respond to sudden traffic flow changes. One type of these parameters in TAC is maximum green time. In this paper, maximum green times are dynamically determined by means of fuzzy control. At first, it monitors the whole intersection traffic conditions and then dynamically adjusts maximum green times. The main advantage of the proposed approach is its capability to adjust maximum green times which are responsive to real time traffic condition in isolated intersections. The efficiency of the proposed method has been evaluated by AIMSUN 7 microscopic traffic simulation tool. The evaluation results show the efficiency and the robustness of the proposed method compared to maximum green times provided by SYNCHRO8.

Optimising a vehicle's approach towards an adaptively controlled intersection

Traffic emissions at controlled intersections can be reduced when the possibilities of infrastructure-to-vehicle communication are put to good use. In this study the authors present two applications: one infrastructure-based and one vehicle-based that in concert are able to significantly reduce the overall emissions at a controlled intersection. The infrastructure-based application employs a model-predictive control approach, an advanced form of traffic-adaptive control. The vehicle-based application uses information gained from the infrastructure-based application (i.e. the estimated time that the vehicle approaching the intersection will be allowed to enter the intersection and cross the stop line) to improve a vehicle's approach towards an intersection. Both applications have the same aim: to avoid unnecessary accelerations, decelerations and delay. For both peak and off-peak hours, the results show that the application of a model-predictive controller as opposed to the more traditional traffic-actuated controller is beneficial both in terms of travel time reduction (∼15.5% in both cases) and carbon-di-oxide (CO2) reduction (2.9 and 9.3%, respectively). Together with an approach advice the amount of CO2 emitted in both cases can be further reduced with an additional 7%, assuming a 100% equipment ratio.

Optimization Method of Intersection Signal Coordinated Control Based on Vehicle Actuated Model

Traditional timing green wave control with predetermined cycle, split, and offset cannot adapt for dynamic real-time traffic flow. This paper proposes a coordinated control method for variable cycle time green wave bandwidth optimization integrated with traffic-actuated control. In the coordinated control, green split is optimized in real time by the measured presence of arriving and/or standing vehicles in each intersection and simultaneously green waves along arterials are guaranteed. Specifically, the dynamic bound of green wave is firstly determined, and then green early-start and green late-start algorithms are presented respectively to accommodate the fluctuations in vehicle arrival rates in each phase. Numerical examples show that the proposed method improves green time, expands green wave bandwidth, and reduces queuing.

Fuzzy Traffic Controller for Urban Intersection Developed in Simulink/SimEvents

This paper presents a traffic signal fuzzy controller for an isolated intersection. The controller is developed based on waiting times and vehicle queues length at shifting traffic light phases. In the first stage, the controller estimates the traffic intensity on each approach. The duration of the green is based on traffic-actuated control. The controller changes traffic light timings and phase sequence to ensure smooth traffic with minimal waiting time, queue length and delay time. We've also made comparison between equal-timing (pre-defined) light control, and fuzzy logic controller - which decides whether to extend or terminate the current signal phase for through movements. The simulation was realized in Simulink/ SimEvents and the results show that the fuzzy approach has significant advantage over the classical pre-timed regime.

Optimization of Control Parameters for Adaptive Traffic-Actuated Signal Control

The authors propose a real-time adaptive signal control model that aims to maintain the adaptive functionality of actuated controllers while improving the performance of traffic-actuated control systems. In this model, a flow-prediction algorithm is formulated to estimate the future vehicle arrival flow for each signal phase at the target intersection on the basis of the available signal-timing data obtained from previous control cycles. Optimal timing parameters are determined on the basis of these estimations and are used as signal-timing data for further optimizations. To be consistent with the operation logic of existing signal-control devices, this model is developed to optimize the basic control parameters that can be found in modern actuated controllers. Microscopic simulation is used to test and evaluate the proposed control model in a calibrated network consisting of 38 actuated signals. Simulation results indicate that this model has the potential to improve the performance of the signalized network under different traffic conditions.

Evaluation and Design of Maximum Green Time Settings for Traffic-Actuated Control

A study is described that was conducted to develop an improved average green time estimation model for traffic-actuated control and to suggest a maximum green time design method that analytically minimizes intersection control delay. Improvements in the green time estimation model include revisions in the concept of additional queue service time, explicit treatment of right turns in lane groups containing both through and right-turning vehicles, and other improvements that include updates based on recent studies and modifications in the approaches taken for the modeling procedure. The proposed maximum green time design procedure consists of four components: ( a) estimation of the average green time of a traffic-actuated phase, ( b) performance evaluation of the system through the 2000 Highway Capacity Manual (HCM) procedure, ( c) formulation of an overall average control delay minimization problem, and ( d) a search process to find the most efficient set of maximum green time parameters that minimize the average control delay at an intersection. Using simulation as a surrogate for field data collection, it was demonstrated that the proposed average green time estimation models offer better results than the one in the 2000 HCM. In addition, it was suggested that on the basis of the improvements demonstrated in terms of design, the proposed maximum green parameter design procedure represents an advancement of the methodology for analysis of signalized intersections.

A rule-based real-time traffic responsive signal control system with transit priority: application to an isolated intersection

Previous experience has shown that real-time, traffic-responsive signal control has the ability to improve traffic operations in urban areas when compared to traditional fixed-time control. However, most of these experiences have focused primarily on the impact on private automobiles. This paper describes the development and evaluation of a fully distributed, real-time, traffic-responsive model named Signal Priority Procedure for Optimization in Real-Time (SPPORT) that explicitly considers the impacts of transit vehicles. This model is unique in two ways. First, the model explicitly considers the interference caused to the general traffic by transit vehicles stopping in the right of way to board and discharge passengers. Second, when considering priority passage for transit vehicles, the potential effects that such preferential treatment might have on other traffic are explicitly quantified. This paper describes the structure of the model and demonstrates its capabilities on an isolated intersection for a range of traffic demands and with and without transit vehicles. The rule-based signal optimization procedure provided delay reductions for most of the traffic conditions examined when compared to both fixed-time and traffic actuated control. While the results reported in this paper are limited to an isolated intersection, the model is capable of being applied to a network of signals. Evaluation of its performance on a corridor has been conducted and will appear in the literature in the near future.

A two-stage fuzzy logic controller for traffic signals

This paper presents the design and evaluation of a fuzzy logic traffic signal controller for an isolated intersection. The controller is designed to be responsive to real-time traffic demands. The fuzzy controller uses vehicle loop detectors, placed upstream of the intersection on each approach, to measure approach flows and estimate queues. These data are used to decide, at regular time intervals, whether to extend or terminate the current signal phase. These decisions are made using a two-stage fuzzy logic procedure. In the first stage, observed approach traffic flows are used to estimate relative traffic intensities in the competing approaches. These traffic intensities are then used in the second stage to determine whether the current signal phase should be extended or terminated. The performance of this controller is compared to that of a traffic-actuated controller for different traffic conditions on a simulated four-approach intersection.

EXPERT SYSTEM OF ACTUATED CONTROL FOR TRAFFIC AT ISOLATED INTERSECTIONS (EXACT): PART III: TESTING AND EVALUATION

Abstract This paper describes the testing stage of the knowledge base of the expert system EXACT (EXpert system of Actuated control for Traffic at isolated intersections). In Tests by two parties (our expert and external experts), more than eighty percent agreement was found between EXACT's solutions and the two parties’ solutions. EXACT was developed to assist traffic engineers in selecting the appropriate type of traffic-actuated controller and its associated features, the type of detector and its configuration, and time settings at a particular intersection. The development of EXACT, including the identification of variables, shell selection, knowledge acquisition, knowledge representation, and knowledgebase development, was presented in earlier papers.

An Integrated Method of Demand Responsive and Traffic Actuated Signal Control

The paper describes the concept of a signal control method that is developed and will be field-tested in the context of the CEC-ATT LLAMD research projects. The method aims to combine the advantages of traditional traffic actuated control and sophisticated demand responsive control in an integrated method. Emphasis is given to the formulation of an objective function which allows for the consideration of any evaluation criterion and particularly for favouring public transport vehicles. The proposed method is based on the use of traditional variables that determine the control. Decomposing the complex problem these variables are assigned to a multi level control strategy.

EXPERT SYSTEM FOR ACTUATED CONTROL OF TRAFFIC AT ISOLATED INTERSECTIONS (EXACT), PART II: THE KNOWLEDGE BASE DEVELOPMENT

Abstract This paper describes the development of the knowledge base of the expert system EXACT (EXpert system of Actuated control for Traffic at isolated intersections). EXACT was developed to assist traffic engineers in selecting the appropriate type of traffic-actuated controller and its associated features, the type of detector and its configuration, and time settings at a particular intersection. The knowledge-base of EXACT consists of three primary elements: frames, parameters, and rules. These elements are activated by a back-word chaining mechanism during a consultation. The ability to explain its reasoning process to the user is one important feature of EXACT. Moreover, EXACT's knowledge-base has the capability of interfacing with graphics program in order to enhance the understandability of the system. A full description of the knowledge-base development of EXACT as well as its graphics interface are presented below.

EXPERT SYSTEM FOR ACTUATED CONTROL OF TRAFFIC AT ISOLATED INTERSECTIONS (EXACT), PART I: KNOWLEDGE ACQUISITION AND REPRESENTATION

Abstract This paper describes the application of expert systems in traffic-actuated control at isolated signalized intersections. The expert system EXACT (EXpert system of ACtuated control for Traffic at isolated intersections) was developed to assist traffic engineers in selecting the appropriate type of traffic-actuated controller and its associated features, the type of detector and its configuration, and time settings at a particular intersection. EXACT was structured to handle both design and operational problems. The first steps of EXACT's development, including variable identification, shell selection, knowledge acquisition, and knowledge representation, are presented in detail in this paper.

Use of Binary Choice Decision Process for Adaptive Signal Control

Adaptive signal control based on a binary-choice decision process for timing adjustment has the potential to replace traffic-actuated control. This emerging control strategy divides time into small intervals. In each interval, a decision is made either to terminate the current green phase or to extend it for another interval. This decision process can produce signal operations that are more efficient than existing traffic-actuated operations. However, there is little understanding of the essential features that should be built into such a process. This paper examines four related issues and gives the results of computer simulation. The analysis reveals that it is sufficient for the binary-choice decision process to rely solely on detected vehicle arrivals for the evaluation of alternative timing decisions. No predictions of other arrivals are needed. It is also found that, for monitoring vehicle arrivals, detectors do not have to be placed at locations more than 8 sec of travel time from the stop line. Furthermore, the binary choice requires the evaluation of an extremely limited number of alternative extensions of the current green phase.

Adaptive Signal Control at Isolated Intersections

The effectiveness of conventional traffic-actuated control deteriorates rapidly as traffic volume increases. This problem can be alleviated through adaptive control. By monitoring vehicular movements an identifying the most likely short-term future flow patterns, adaptive control adjusts signal timings in an attempt to optimize signal operations. Both field test and simulation analysis have shown that adaptive control has the potential to provide improved signal operations. Further research is needed to develop a versatile and effective adaptive control logic that can replace traffic-actuated logics. The primary objective of this paper is to discuss the issues and research needs concerning the development of adaptive control logics for applications at isolated intersections. The discussions are focused on the information needs for adaptive control, the selection between a binary choice process and a sequencing process for timing adjustment, and future research needs. To be widely applicable, an adaptive control logic needs a vehicle-monitoring system that can provide reliable information and facilitate simulataneous use of real time and advance information for decision making.

A General Functional Structure for Traffic Actuated Local Controllers

A hierarchical multilayer control structure is proposed for traffic-actuated controllers of isolated intersections, which decomposes vertically the control problem by considering at each layer a subproblem with relaxed constraints qualified by its upper layer subproblem solution. A very elementary implementation is then described, and its performance is tested by simulation in a quite complex situation, and is compared with a conventional traffic-actuated control strategy operating on a predetermined phase basis. The results show drastic reductions for the average stop time, which range from 14% to over 30% in steady traffic conditions.