Real-time Dynamic Traffic Assignment Research Articles

Estimation and prediction of the OD matrix in uncongested urban road network based on traffic flows using deep learning

In this article, we propose a new method for OD (Origin–Destination)​ matrix prediction based on traffic data using deep learning. The input values of the developed model were determined based on data on the structure of the road network, origin and destination points of trips, as well as data on traffic intensity on road network sections recorded by video-sensing devices. The advantage of the method is that the complex process of data acquisition and processing is not required for the estimation and prediction of the matrix. Historical data and the iterative method of estimating a prior OD matrix were used only to generate training sequences for the neural network. The proposed method using deep learning neural networks with the long short-term memory (LSTM) or autoencoders layers (DLNa — deep learning networks with autoencoders) is characterized by relatively high accuracy and resistance to temporary missing data from several measurement points located in the urban road network. The case study was conducted for a network of a medium-sized city in Poland. The results show (average MAPE = 7.18% (LSTM), 6.80% (DLNa)) that the proposed method can have a practical implementation in real-time dynamic traffic assignment (DTA) systems for ITS applications. The proposed method of short-term forecasting the OD matrix does not require questionnaire research or detailed information on spatial development. Therefore, it is not as expensive and time-consuming as the methods based on these data.

Day-to-Day Learning Framework for Online Origin–Destination Demand Estimation and Network State Prediction

Origin–destination (O–D) demand is a critical component in both online and offline dynamic traffic assignment (DTA) systems. Recent advances in real-time DTA applications in large networks call for robust and efficient methodologies for online O–D demand estimation and prediction. This study presents a day-to-day learning framework for a priori O–D demand, along with a predictive data-driven O–D correction approach for online consistency between predicted and observed (sensor) values. When deviations between simulation and real world are observed, a consistency-checking module initiates O–D demand correction for the given prediction horizon. Two predictive correction methods are suggested: 1) simple gradient method, and 2) Taylor approximation method. New O–D demand matrices, corrected for 24 simulation hours by the correction module, are used as the updated a priori demand for the next day simulation. The methodology is tested in a real-world network, Kansas City, MO, for a 3-day period. Actual tests in real-world networks of online DTA systems have been very limited in the literature and in actual practice. The test results are analyzed in time and space dimensions. The overall performance of observed links is assessed. To measure the impact of O–D correction and daily O–D updates, traffic prediction performance with the new modules is compared with the base case. Predictive O–D correction improves prediction performance in a long prediction window. Also, daily updated O–D demand provides better initial states for traffic prediction, enhancing prediction in short prediction windows. The two modules collectively improve traffic prediction performance of the real-time DTA system.

Improving Scalability of Generic Online Calibration for Real-Time Dynamic Traffic Assignment Systems

Flexible calibration of dynamic traffic assignment (DTA) systems in real time has important applications in effective traffic management. However, the existing approaches are either limited to small networks or to a specific class of parameters. In this light, this study presents a framework to systematically reduce the dimension of the generic online calibration problem, making it more scalable. Specifically, a state–space formulation of the problem in the reduced dimension space is proposed. Following this the problem is solved using the constrained extended Kalman filter, which is made tractable because of the low dimensionality of the formulated problem. The effectiveness of the proposed approach is demonstrated using a real-world network leading to better state estimation by 13% and better state predictions by 11%—with a 50 fold dimensionality reduction. Insights into choosing the right degree of dimensionality reduction are also discussed. This work has the potential for a more widespread application of real-time DTA systems in practice.

PATH PRIORITIZATION AND OPERATIONAL CONSISTENCY FOR DYNAMIC TRAFFIC ASSIGNMENT: AN ANALYTIC HIERARCHICAL APPROACH

An operational consistency model for real-time dynamic traffic assignment (DTA) applications seeks to correct the time-dependent path assignment within a rolling horizon scheme. This study extends an existing consistency framework to develop a hierarchy for the time-dependent path set based upon their relative importance to ensuring consistency. Using the analytic hierarchy process, the eigenvalue associated with a path is identified as the parameter which enables the rank ordering of paths. The ability to identify a subset of dominant paths relative to enhancing consistency enhances the computational viability of the consistency framework for real-time implementation and has significant practical implications. Additionally, it provides insights on the complex dynamics that are inherent to the operational consistency problem.

A structural state space model for real-time traffic origin–destination demand estimation and prediction in a day-to-day learning framework

Dynamic origin–destination (OD) estimation and prediction is an essential support function for real-time dynamic traffic assignment model systems for ITS applications. This paper presents a structural state space model to systematically incorporate regular demand pattern information, structural deviations and random fluctuations. By considering demand deviations from the a priori estimate of the regular pattern as a time-varying process with smooth trend, a polynomial trend filter is developed to capture possible structural deviations in real-time demand. Based on a Kalman filtering framework, an optimal adaptive procedure is further proposed to capture day-to-day demand evolution, and update the a priori regular demand pattern estimate using new real-time estimates and observations obtained every day. These models can be naturally integrated into a real-time dynamic traffic assignment system and provide an effective and efficient approach to utilize the real-time traffic data continuously in operational settings. A case study based on the Irvine test bed network is conducted to illustrate the proposed methodology.

Number and Location of Sensors for Real-Time Network Traffic Estimation and Prediction: Sensitivity Analysis

Installing and maintaining sensors in a transportation network can be expensive. The motivation for this research is finding the best way to deploy finite resources and generate a network detection system in a manner that produces minimal estimation errors. The analysis uses a simulation-based real-time network traffic estimation and prediction system based on dynamic traffic assignment (DTA) methodology to analyze different levels of detection and different sensor locations in a portion of the Chesapeake Highway Advisories Routing Traffic (CHART) network (between Washington, D.C., and Baltimore, Maryland). This study provides a conceptual framework of the sensor location problem and a theoretical description of the objectives associated with the sensor location problem. A sensitivity analysis of the estimation and prediction quality with the DYNASMART-X real-time DTA system in relation to sensor number and location is conducted in the Maryland CHART network. The analysis considers both randomly generated location scenarios and scenarios based on engineering judgment. The analysis reveals the importance of providing detection in specific locations of the network and the dependence of the value of additional detection on the specific location selected.

Number and Location of Sensors for Real-Time Network Traffic Estimation and Prediction

Installing and maintaining sensors in a transportation network can be expensive. The motivation for this research is finding the best way to deploy finite resources and generate a network detection system in a manner that produces minimal estimation errors. The analysis uses a simulation-based real-time network traffic estimation and prediction system based on dynamic traffic assignment (DTA) methodology to analyze different levels of detection and different sensor locations in a portion of the Chesapeake Highway Advisories Routing Traffic (CHART) network (between Washington, D.C., and Baltimore, Maryland). This study provides a conceptual framework of the sensor location problem and a theoretical description of the objectives associated with the sensor location problem. A sensitivity analysis of the estimation and prediction quality with the DYNASMART-X real-time DTA system in relation to sensor number and location is conducted in the Maryland CHART network. The analysis considers both randomly generated location scenarios and scenarios based on engineering judgment. The analysis reveals the importance of providing detection in specific locations of the network and the dependence of the value of additional detection on the specific location selected.

Online Consistency Checking and Origin-Destination Demand Updating: Recursive Approaches with Real-Time Dynamic Traffic Assignment Operator

To maintain an internally consistent representation with actual traffic conditions, this paper presents an origin-destination (O-D) demand consistency-checking and updating model for online dynamic traffic assignment operation. Both predictive and reactive approaches are proposed to minimize the deviations between simulated states and real-world observations and O-D demand adjustment magnitude. The two objectives are combined into a weighted linear quadratic function to construct a guaranteed overdetermined optimization problem. Alternative recursive solution algorithms are presented to design an efficient feedback controller that regulates the demand input for the real-time dynamic traffic assignment simulator. The proposed model is tested with field data from the Irvine test bed network.

Online Consistency Checking and Origin–Destination Demand Updating

To maintain an internally consistent representation with actual traffic conditions, this paper presents an origin–destination (O-D) demand consistency-checking and updating model for online dynamic traffic assignment operation. Both predictive and reactive approaches are proposed to minimize the deviations between simulated states and real-world observations and O-D demand adjustment magnitude. The two objectives are combined into a weighted linear quadratic function to construct a guaranteed overdetermined optimization problem. Alternative recursive solution algorithms are presented to design an efficient feedback controller that regulates the demand input for the real-time dynamic traffic assignment simulator. The proposed model is tested with field data from the Irvine test bed network.

Adaptive Speed Estimation Using Transfer Function Models for Real-Time Dynamic Traffic Assignment Operation

The application of a transfer function model (TFM) in a real-time dynamic traffic assignment system is investigated. The motivation is to improve the speed estimation method to enable better system consistency with reality in real-time operation. The study is conducted by adopting a TFM derived from actual detector data in San Antonio, Texas. This model is then used in the traffic simulation module of the DYNASMART-X dynamic traffic assignment system to update the network link speeds. A nonlinear least-squares optimization algorithm, implemented for this study, is coupled with DYNASMART-X to enable adaptive estimation of the TFM parameters. Simulation-based experiments are carried out on the Fort Worth test network. These experiments are designed to evaluate the TFM performance and to gain insight into its operational properties under different conditions. The results show that the TFM, both adaptive and nonadaptive, can consistently approximate the true underlying speed-density dynamics. Of significant importance is the transfer-ability and robustness of TFMs in different settings. The outcome of this research substantiates the premise that good speed estimation can be achieved through the use of TFMs.

Laboratory Evaluation of Real-Time Dynamic Traffic Assignment Systems

The success of Advanced Traffic Management Systems (ATMS) and Advanced Traveler Information Systems (ATIS) depends on the availability and dissemination of timely and accurate estimates of prevailing and emerging traffic conditions, as well as the provision of faithful and coherent pretrip and en-route travel information and advisory to travelers. Real-time Dynamic Traffic Assignment (DTA) systems, envisioned as an ATMS support system that resides in transportation management centers, are being developed to supply these estimates and information. To ensure the deployment potential of such DTA systems in the field, extensive laboratory evaluations are required throughout the development process. This article discusses a software environment that economically supports the development and evaluation of DTA systems. A major simulation component developed under this environment is called a DTA evaluation system (DES). The DES is a virtual representation of a transportation system in which ATMS and ATIS technologies are deployed. It simulates the trip-making and vehicle-driving behaviors of travelers in response to real-time information and advisories from ATIS and ATMS. This article presents (1) major DES modeling capabilities required for developing and evaluating DTA systems, (2) a modular and extensible modeling framework for the DES, and (3) a top-level, object-oriented, and distributed DES design.

Real-Time Dynamic Traffic Assignment and Path-Based Signal Coordination; Application to Network Traffic Management

As an example of integrating signal control with network traffic assignment, the notion of path-based signal coordination in transportation networks is introduced and illustrated. The use of a real-time dynamic traffic assignment system allows prediction of the traffic flow patterns and identification of the dominant paths in the network. The signalized intersections along these paths, which may consist of combinations of straight sections and turning movements, are coordinated to increase the capacity of the freeway and the surface street system to efficiently absorb diverted traffic from the freeway. A set of experiments is designed to compare the network performance under the path-based coordination scheme with no coordination and arterial-based coordination. The experiments are conducted for both pretimed and vehicle-actuated signal control. An actual network, which represents the south-central part of the Fort Worth area, is used in these experiments. The network consists of a freeway (I-35W) surrounded by a street network with a total of 178 nodes and 441 arcs. The path-based coordination scheme is shown to outperform arterial-based coordination in the cases of both pretimed fixed control and vehicle-actuated signal control. A significant reduction in the network average travel time under the path-based coordination scheme is observed compared with the other schemes. This application further illustrates the benefits of real-time dynamic traffic assignment for advanced traffic management under incident conditions.

On-Line Monitoring System for Real-Time Traffic Management Applications

This paper is concerned with the problem of on-line monitoring of a traffic network. The problem was motivated by the need of real-time dynamic traffic assignment systems to maintain an internal representation of the traffic network consistent with that of the actual network. Potential error sources that can cause inconsistencies between the internal and the actual state of the network are identified and classified. The issues of observability and state variable definition of traffic networks are discussed. A framework for a monitoring system is sketched that aims to maintain a representation of the state of a network consistent with that of the actual network by applying on-line and off-line adjustments. Some solution approaches and computational results are presented for the main modules of the framework.

Real-Time Dynamic Traffic Assignment for Route Guidance: Comparison of Global Predictive vs. Local Reactive Strategies Under Stochastic Demands

This paper describes two strategies for real-time route guidance of equipped motorists in a congested traffic network, and presents a comparative assessment of the strategies’ effectiveness and robustness under stochastic origin-destination (O-D) demands. The first approach consists of a global network-level system optimal assignment algorithm implemented in a rolling-horizon framework, with predicted O-D demands. The second consists of simpler locally-oriented rules that react to measured conditions (no prediction). Results of numerical experiments using a simulation-assignment model (DYNASMART) under different scenarios of demand variability and forecast error are presented with regard to (1) sensitivity of the RH solution to stage length, roll period and forecast errors, and (2) comparative performance and robustness of the two approaches under stochastically varying demands and imperfect forecasts.