Random Process Model for Urban Traffic Flow Using a Wavelet‐Bayesian Hierarchical Technique

Abstract: The existing well‐known short‐term traffic forecasting algorithms require large traffic flow data sets, including information on current traffic scenarios to predict the future traffic conditions. This article proposes a random process traffic volume model that enables estimation and prediction of traffic volume at sites where such large and continuous data sets of traffic condition related information are unavailable. The proposed model is based on a combination of wavelet analysis (WA) and Bayesian hierarchical methodology (BHM). The average daily “trend” of urban traffic flow observations can be reliably modeled using discrete WA. The remaining fluctuating parts of the traffic volume observations are modeled using BHM. This BHM modeling considers that the variance of the urban traffic flow observations from an intersection vary with the time‐of‐the‐day. A case study has been performed at two busy junctions at the city‐centre of Dublin to validate the effectiveness of the strategy.     

Estimating dynamic origin–destination demand: A hybrid framework using license plate recognition data

This article proposes a hybrid framework for estimating dynamic origin–destination (OD) demand that fully exploits the information available in license plate recognition (LPR) data. A Bayesian path reconstruction model is initially developed to replenish the lost information resulting from the recognition error and insufficient coverage rate of the LPR system. The link flows, initial OD demand, left‐turning flows, and partial path flows are derived based on the reconstructed data. Subsequently, with the information derived, a two‐step ordinary least squares (OLS) OD estimation model is formulated, which incorporates the output from the Bayesian model and coestimates the OD demand and assignment matrix. The proposed framework is qualitatively validated using the real‐world LPR data collected from Langfang City, Hebei Province, China, and is quantitatively validated using the synthesized simulation data for the simplified road network of Langfang. The results show that the proposed model can estimate OD demand distribution with a mean absolute percentage error (MAPE) of about 30%. We also tested the model with different LPR coverage rates, with results showing that an LPR coverage rate of over 50% is required to obtain reasonable results.     

A graph deep learning method for short‐term traffic forecasting on large road networks

Short‐term traffic flow prediction on a large‐scale road network is challenging due to the complex spatial–temporal dependencies, the directed network topology, and the high computational cost. To address the challenges, this article develops a graph deep learning framework to predict large‐scale network traffic flow with high accuracy and efficiency. Specifically, we model the dynamics of the traffic flow on a road network as an irreducible and aperiodic Markov chain on a directed graph. Based on the representation, a novel spatial–temporal graph inception residual network (STGI‐ResNet) is developed for network‐based traffic prediction. This model integrates multiple spatial–temporal graph convolution (STGC) operators, residual learning, and the inception structure. The proposed STGC operators can adaptively extract spatial–temporal features from multiple traffic periodicities while preserving the topology information of the road network. The proposed STGI‐ResNet inherits the advantages of residual learning and inception structure to improve prediction accuracy, accelerate the model training process, and reduce difficult parameter tuning efforts. The computational complexity is linearly related to the number of road links, which enables citywide short‐term traffic prediction. Experiments using a car‐hailing traffic data set at 10‐, 30‐, and 60‐min intervals for a large road network in a Chinese city shows that the proposed model outperformed various state‐of‐the‐art baselines for short‐term network traffic flow prediction.     

Two‐Dimensional Simulation of Turning Behavior in Potential Conflict Area of Mixed‐Flow Intersections

The potential conflict area of intersection is the space where conflicting traffic flows pass through in the same signal phase. At this area, turning vehicles interact with most traffic flows, which introduce complex features including variation of trajectories and shared‐priority phenomenon. The traditional one‐dimensional simulation oversimplifies these features with lane‐based assumption. This study integrates the modified social force model with behavior decision and movement constraints to reproduce the two‐dimensional turning process. The method is framed into a three‐layered mathematical model. First, the decision layer dynamically makes decision for turning patterns. Then the operation layer uses the modified social force model to initially generate vehicle movements. Finally, the constraint layer modifies the vehicular motion with vehicle dynamics constraints, boundary of intersection and the collision avoidance rule. The proposed model is validated using trajectories of left‐turn vehicles at a real‐world mixed‐flow intersection with nonprotected signal phases, resulting in a more realistic simulation than previous methods. The distributions of decision points and travel time in simulation are compared with the empirical data in statistics. Moreover, the spatial distribution of simulated trajectories is also satisfactory.     

Allocation of Link Flow Detectors for Origin‐Destination Matrix Estimation—A Comparative Study

Abstract: Origin‐destination (OD) matrices are essential for various analyses in the field of traffic planning, and they are often estimated from link flow observations. We compare methods for allocating link flow detectors to a traffic network with respect to the quality of the estimated OD‐matrix. First, an overview of allocation methods proposed in the literature is presented. Second, we construct a controlled experimental environment where any allocation method can be evaluated, and compared to others, in terms of the quality of the estimated OD‐matrix. Third, this environment is used to evaluate and compare three fundamental allocation methods. Studies are made on the Sioux Falls network and on a network modeling the city of Linköping. Our conclusion is, that the most commonly studied approach for detector allocation, maximizing the coverage of OD‐pairs, seems to be unfavorable for the quality of the estimated OD‐matrix.     

An Approach to Dynamic Classification of Traffic Flow Patterns

Abstract: Analyses on the dynamics of traffic flow, ranging from intersection flows to network‐wide flow propagation, require accurate information on time‐varying local traffic flows. To effectively determine the flow performance measures and consequently the congestion indicators of segmented road pieces, the ability to process such data in real time is out of the question. In this article, a dynamic approach to specify flow pattern variations is proposed mainly concentrating on the incorporation of neural network theory to provide real‐time mapping for traffic density simultaneously in conjunction with a macroscopic traffic flow model. To deal with the noise and the wide scatter of raw flow measures, a filtering is applied prior to modeling processes. Filtered data are dynamically and simultaneously input to processes of neural density mapping and traffic flow modeling. The classification of flow patterns over the fundamental diagram, which is dynamically plotted with the outputs of the flow modeling subprocess, is obtained by considering the density measure as a pattern indicator. Densities are mapped by selected neural approximation method for each simulation time step considering explicitly the flow conservation principle. Simultaneously, mapped densities are matched over the fundamental diagram to specify the current corresponding flow pattern. The approach is promising in capturing sudden changes on flow patterns and is open to be utilized within a series of intelligent management strategies including especially nonrecurrent congestion effect detection and control.     

基于交叉路口的动态OD反推模型与算法研究

交叉路口各进出口道之间的实时交通量是信号控制系统重要的输入数据 ,也是难以获得的数据。本文回顾了动态OD反推理论的发展历程 ,以交叉路口各进出口道检测到的路段流量时间序列为基础 ,分析了交叉路口实时交通量的特点 ,提出了基于交叉路口的改进参数优化模型 ,设计了遗传算法对问题进行求解 ,并讨论了算法中的六个关键问题。实例研究表明 ,模型和算法具有较高的效率和准确性 ,能够应用于在线系统。     

Extensible prototype learning for real-time traffic signal control

Congestion resolution continues to remain a challenge even though various signal control systems have been developed for traffic-intersection control. To address this issue, reinforcement learning (RL)-based approaches that focus on solving the associated data-driven problems have been proposed. However, only a few methods have been developed and applied to dual-ring traffic signal control systems. Therefore, we develop an RL-based traffic signal control model for such a system to efficiently allocate the green interval in different oversaturation states of the conflicting phases. The proposed model employs a deep deterministic policy gradient algorithm to optimize the green value in the continuous action space. Further, we develop an extensible prototype learning framework for application to new intersections without additional transfer learning. The proposed model is validated based on morning peak hours in a simulation environment that reflects the actual intersection phase system and minimum green time constraints. The proposed model achieves an average 20% intersection delay reduction, compared with the fixed control method.     

Traffic Estimation and Optimal Counting Location Without Path Enumeration Using Bayesian Networks

Abstract:   This article deals with the problem of estimating and updating the origin-destination matrix and link flows from traffic counts and its optimal location. A combination (bi-level) of an OD-pair matrix estimation model based on Bayesian networks, and a Wardrop-minimum-variance model, which identifies origins and destinations of link flows, is used to estimate OD-pair and unobserved link flows based on some observations of links and/or OD-pair flows. The Bayesian network model is also used to select the optimal number and locations of the links counters based on maximum correlation. Finally, the proposed methods are illustrated by their application to the Nguyen–Dupuis and the Ciudad Real networks.     

Spatial dependence in (origin‐destination) air passenger flows

We explore the estimation of origin‐destination (OD), city‐pair, air passenger flows. Our dataset contains 279 cities, worldwide, over 2010‐2012. Allowing for two gravity model specifications (log‐normal and Poisson), we compare non‐spatial and spatial models. We are the first to apply spatial econometric flow models and eigenfunction spatial filtering approaches to air transport. Distinguishing between origin, destination and network effects, we determine the impact and significance of a change in a variable in a given city, on flows originating from and going to that city and originating from a different city and going to an alternative city. Finally, we compare models based on different specifications of the spatial weight matrix.     

Graph‐Based Modeling of Information Flow Evolution and Propagation under V2V Communications‐Based Advanced Traveler Information Systems

A vehicle equipped with a vehicle‐to‐vehicle (V2V) communications capability can continuously update its knowledge on traffic conditions using its own experience and anonymously obtained travel experience data from other such equipped vehicles without any central coordination. In such a V2V communications‐based advanced traveler information system (ATIS), the dynamics of traffic flow and intervehicle communication lead to the time‐dependent vehicle knowledge on the traffic network conditions. In this context, this study proposes a graph‐based multilayer network framework to model the V2V‐based ATIS as a complex system which is composed of three coupled network layers: a physical traffic flow network, and virtual intervehicle communication and information flow networks. To determine the occurrence of V2V communication, the intervehicle communication layer is first constructed using the time‐dependent locations of vehicles in the traffic flow layer and intervehicle communication‐related constraints. Then an information flow network is constructed based on events in the traffic and intervehicle communication networks. The graph structure of this information flow network enables the efficient tracking of the time‐dependent vehicle knowledge of the traffic network conditions using a simple graph‐based reverse search algorithm and the storage of the information flow network as a single graph database. Further, the proposed framework provides a retrospective modeling capability to articulate explicitly how information flow evolves and propagates. These capabilities are critical to develop strategies for the rapid flow of useful information and traffic routing to enhance network performance. It also serves as a basic building block for the design of V2V‐based route guidance strategies to manage traffic conditions in congested networks. Synthetic experiments are used to compare the graph‐based approach to a simulation‐based approach, and illustrate both memory usage and computational time efficiencies.     

Spatiotemporal gated graph attention network for urban traffic flow prediction based on license plate recognition data

The accurate forecasting of traffic states is an essential application of intelligent transportation system. Due to the periodic signal control at intersections, the traffic flow in an urban road network is often disturbed and expresses intermittent features. This study proposes a forecasting framework named the spatiotemporal gated graph attention network (STGGAT) model to achieve accurate predictions for network-scale traffic flows on urban roads. Based on license plate recognition (LPR) records, the average travel times and volume transition relationships are estimated to construct weighted directed graphs. The proposed STGGAT model integrates a gated recurrent unit layer, a graph attention network layer with edge features, a gated mechanism based on the bidirectional long short-term memory and a residual structure to extract the spatiotemporal dependencies of the approach- and lane-level traffic volumes. Validated on the LPR system in Changsha, China, STGGAT demonstrates superior accuracy and stability to those of the baselines and reveals its inductive learning and fault tolerance capabilities.     

Real-Time OD Estimation Using Automatic Vehicle Identification and Traffic Count Data

A key input to many advanced traffic management operations strategies are origin–destination (OD) matricies. In order to examine the possibility of estimating OD matricies in real-time, two constrained OD estimators, based on generalized least squares and Kalman filtering, were developed and tested. A one-at-a-time processing method was introduced to provide an efficient organized framework for incorporating observations from multiple data sources in real-time. The estimators were tested under different conditions based on the type of prior OD information available, the type of assignment available, and the type of link volume model used. The performance of the Kalman filter estimators also was compared to that of the generalized least squares estimator to provide insight regarding their performance characteristics relative to one another for given scenarios. Automatic vehicle identification (AVI) tag counts were used so that observed and estimated OD parameters could be compared. While the approach was motivated using AVI data, the methodology can be generalized to any situation where traffic counts are available and origin volumes can be estimated reliably. The primary means by which AVI data was utilized was through the incorporation of prior observed OD information as measurements, the inclusion of a deterministic link volume component that makes use of OD data extracted from the latest time interval from which all trips have been completed, and through the use of link choice proportions estimated based on link travel time data. It was found that utilizing prior observed OD data along with link counts improves estimator accuracy relative to OD estimation based exclusively on link counts.     

Simulation model based on Monte Carlo method for traffic assignment in local area road network

For a local area road network, the available traffic data of traveling are the flow volumes in the key intersections, not the complete OD matrix. Considering the circumstance characteristic and the data availability of a local area road network, a new model for traffic assignment based on Monte Carlo simulation of intersection turning movement is provided in this paper. For good stability in temporal sequence, turning ratio is adopted as the important parameter of this model. The formulation for local area road network assignment problems is proposed on the assumption of random turning behavior. The traffic assignment model based on the Monte Carlo method has been used in traffic analysis for an actual urban road network. The results comparing surveying traffic flow data and determining flow data by the previous model verify the applicability and validity of the proposed methodology.     

Kernel Smoothing Method Applicable to the Dynamic Calibration of Traffic Flow Models

Abstract: Real time traffic flow simulation models are used to provide traffic information for dynamic traffic management systems. Those simulation models are supplied by traffic data in order to estimate and predict traffic conditions in unobserved sections of a traffic network. In general, most of recent real time traffic simulators are based on the macroscopic model because the macroscopic model replicates the average traffic behavior in terms of observable variables such as (time–space) flow and speed at a relatively fast computational time. Like other simulation models, an important aspect of the real time macroscopic simulator is to calibrate the model parameters online. The most conventional way of the online calibration is to add a random walk to the parameters to constitute an augmentation of the traffic variables and the model parameters to be estimated. Actually, this method allows the parameters to vary at every time step and, therefore, describes the adaptation of the model to the prevailing traffic conditions. However, it has been reported that the use of the random walk results in a loss of information and an increase of the covariance of parameters, which consequently leads to posteriors that are far more diffuse than the theoretical posteriors for the true parameters. To this end, this article puts forward a Kernel density estimation technique in the calibration process to handle the covariance issue and to avoid the information loss. The Kernel density estimation technique is embedded in the particle filter algorithm, which is extended to the calibration problems. The proposed framework is investigated using real‐life data collected in a freeway in England.     

On the Probabilistic and Physical Consistency of Traffic Random Variables and Models

In this article we deal with the probabilistic and physical consistency of traffic‐related random variables and models. We analyze and discuss the conditions for a model to be consistent from two different points of view: probabilistic and physical (dimensional analysis). The first, leads us to the concept of stability in general and reproductivity in particular because, for example, origin–destination (OD) and link flows are the sum of route flows and route travel times are the sum of link travel times. This implies stability with respect to sums (reproductivity). Normal models are justified because when the number of summands increases the averages approach the normal distribution. Similarly, stability with respect to minimum or maximum operations arises in practice. From the dimensional analysis point of view, some models are demonstrated not to be convenient. In particular, it is shown that some families of distributions are valid only for dimensionless variables. All these and other problems are discussed and some proposed models in the literature are analyzed from these two points of view. When some families fail to satisfy the desired properties, alternative models are provided via extension of the original families. Finally, some simple examples and conclusions are given to summarize the analysis.     

A network traffic assignment model for autonomous vehicles with parking choices

This article is the first in the literature to investigate the network traffic equilibrium for traveling and parking with autonomous vehicles (AVs) under a fully automated traffic environment. Given that AVs can drop off the travelers at their destinations and then drive to the parking spaces by themselves, we introduce the joint equilibrium of AV route choice and parking location choice, and develop a variational inequality (VI)‐based formulation for the proposed equilibrium. We prove the equivalence between the proposed VI model and the defined equilibrium conditions. We also show that the link flow solution at equilibrium is unique, even though both the route choices and parking choices are endogenous when human‐occupied AV trips (from origin to destination) and empty AV trips (from destination to parking) are interacting with each other on the same network. We then develop a solution methodology based on the parking‐route choice structure, where we adjust parking choices in the upper level and route choices in the lower level. Numerical analysis is conducted to explore insights from the introduced modeling framework for AV network equilibrium. The results reveal the significant difference in network equilibrium flows between the AV and non‐AV situations. The results also indicate the sensitivity of the AV traffic pattern to different factors, such as value of time, parking pricing, and supply. The proposed approach provides a critical modeling device for studying the traffic equilibrium under AV behavior patterns, which can be used for the assessment of parking policies and infrastructure development in the future era of AVs.     

Network‐wide traffic speed forecasting: 3D convolutional neural network with ensemble empirical mode decomposition

Accurate traffic speed forecasting is one of the most critical tasks in proactive traffic management and the deployment of advanced traveler information systems. This paper proposes a hybrid forecasting approach named DeepEnsemble by integrating the three‐dimensional convolutional neural network (3D CNN) with ensemble empirical mode decomposition (EEMD). There are four steps in this hybrid approach. First, EEMD is adopted to decompose the complex traffic speed time series data with noise into several intrinsic mode functions (IMFs) and a residue. Second, a three‐dimensional tensor is established and fed into 3D CNN for prediction. Third, the output of 3D CNN prediction is obtained by a linear combination of the results of all components. Finally, the 3D CNN prediction output, external features, and historical features are fused to predict the network‐wide traffic speed simultaneously. The proposed DeepEnsemble approach is tested on the three‐month traffic speed series data of a real‐world large‐scale urban expressway network with 308 traffic flow detectors in Beijing, China. The experimental results indicate that DeepEnsemble outperforms the state‐of‐the‐art network‐wide traffic speed forecasting models. 3D CNN learns temporal, spatial, and depth information better than 2D CNN. Moreover, forecasting accuracy can be improved by employing EEMD. DeepEnsemble is a promising model with scalability and portability for network‐wide traffic speed prediction and can be further extended to conduct traffic status monitoring and congestion mitigation strategies.     

Wavelet‐Based Denoising for Traffic Volume Time Series Forecasting with Self‐Organizing Neural Networks

Abstract: In their goal to effectively manage the use of existing infrastructures, intelligent transportation systems require precise forecasting of near‐term traffic volumes to feed real‐time analytical models and traffic surveillance tools that alert of network links reaching their capacity. This article proposes a new methodological approach for short‐term predictions of time series of volume data at isolated cross sections. The originality in the computational modeling stems from the fit of threshold values used in the stationary wavelet‐based denoising process applied on the time series, and from the determination of patterns that characterize the evolution of its samples over a fixed prediction horizon. A self‐organizing fuzzy neural network is optimized in its configuration parameters for learning and recognition of these patterns. Four real‐world data sets from three interstate roads are considered for evaluating the performance of the proposed model. A quantitative comparison made with the results obtained by four other relevant prediction models shows a favorable outcome.     

Robust Optimal Lane Allocation for Isolated Intersections

Lane allocation including approach and exit lane numbers and lane markings of approach lanes plays an important role in improving the capacity of an intersection. Conventional approaches for optimizing lane allocation often ignore fluctuations in traffic demand (TD). This article presents a stochastic model for robust optimal lane allocation of an isolated intersection under stochastic traffic conditions. This model is built in three steps. In the first step, an enhanced lane‐based model in the form of a binary mixed‐integer nonlinear program is proposed to optimize lane allocation and traffic signals for both vehicles and pedestrians in a unified framework under deterministic traffic conditions. In the second step, a two‐level stochastic model is developed to obtain a robust lane allocation that is less sensitive to traffic flow fluctuations considering the flexibility of traffic signals. In the third step, the two‐level model is further transformed into a TD‐based stochastic model in a two‐phase form to reduce the solution dimension for efficient computation. A TD‐based genetic algorithm procedure is presented for solvability. Numerical studies are conducted to validate the model formulations and solution algorithms.