Advanced Traffic Information Systems Research Articles

An empirical investigation of driver car-following risk evolution using naturistic driving data and random parameters multinomial logit model with heterogeneity in means and variances

This study aims to address the questions of how driving risk evolves during car-following processes and what factors contribute to the underlying evolution patterns. An empirical study is conducted using real world car-following data collected in the Shanghai Naturalistic Driving Study (SH-NDS). The evolution of the driving risk induced by the dynamic coupling between the leading and following vehicles during the car-following process is characterized by how an instantaneous crash-risk measure - rear crash risk index (RCRI) - changes by time. A spectral clustering analysis is first conducted to classify the driving risk evolution of the observed car-following maneuvers, showing the existence of five distinctive risk evolution patterns in the car-following processes. In order to investigate the relationship between the identified driving risk evolution clusters and their contributing factors, a regression analysis employing a random parameter multinomial logit model with heterogeneity in means and variances is followed, revealing several significant contributing factors to the car-following risk evolution patterns, such as congestion level, driver’s ability to maintain stable headways, and vehicle deceleration. This study has provided important insights into driving risk from the new perspective of risk evolution patterns, which is expected to have significant implications for the future development of advanced traffic management and traveler information systems (ATMS/ATIS) strategies, advanced driver assistance systems (ADAS), and connected and autonomous vehicles (CAV).

Spatio-Temporal vehicle traffic flow prediction using multivariate CNN and LSTM model

Traffic congestion is a major problem in developing and developed countries vehicle traffic management systems. Traffic control system works based on the idea of removing instabilities and avoid accidents in order to minimize the traffic and maximize the vehicle flow. To control the congestion need to predict the upcoming traffic flow and it will be useful for Advanced Traffic Information Systems (ATIS), Advanced Traffic Management Systems (ATMS) and traffic analytics. Non–linear historical data and uncertain factors influence the vehicle congestion at peak hours which cannot be considered in existing algorithms. This study proposes hybrid neural network algorithms such as Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) network for short term traffic flow prediction based on multivariate analysis. Widely referred datasets Performance Measurement Systems (PEMS) and Mesowest have been used to evaluate this model. Experiment results shows that CNN-LSTM Hybrid prediction model achieves high accuracy compared with other models.

A GATs-GAN framework for road traffic states forecasting

Short-term traffic states forecasting of road networks based on real-time data is an important component of intelligent transportation systems, especially advanced traffic management systems and traveller information systems. By considering the influence of both space and time dimensions, we proposed a novel GATs-GAN framework for the forecasting of traffic states. First, to capture spatial traffic relationships, the traffic topological graph network is set up based on the connection of traffic sections. Then, the first-order neighbours and high-order neighbours of traffic networks can be structured. Graph attention networks (GATs) are used to obtain the hidden features of input traffic data by training the attention between nodes in high-order neighbours. Based on two traffic networks in California and Seattle in the United States, we find that the GATs-GAN with high-order neighbours can satisfactorily estimate the traffic data and performs better than the baseline methods and comparative experiments.

Highway Travel Time Prediction of Segments Based on ANPR Data considering Traffic Diversion

Travel time is one of the most critical parameters in proactive traffic management and the deployment of advanced traveler information systems. This paper proposes a hybrid model named LSTM-CNN for predicting the travel time of highways by integrating the long short-term memory (LSTM) and the convolutional neural networks (CNNs) with the attention mechanism and the residual network. The highway is divided into multiple segments by considering the traffic diversion and the relative location of automatic number plate recognition (ANPR). There are four steps in this hybrid approach. First, the average travel time of each segment in each interval is calculated from ANPR and fed into LSTM in the form of a multidimensional array. Second, the attention mechanism is adopted to combine the hidden layer of LSTM with dynamic temporal weights. Third, the residual network is introduced to increase the network depth and overcome the vanishing gradient problem, which consists of three pairs of one-dimensional convolutional layers (Conv1D) and batch normalization (BatchNorm) with the rectified linear unit (ReLU) as the activation function. Finally, a series of Conv1D layers is connected to extract features further and reduce dimensionality. The proposed LSTM-CNN approach is tested on the three-month ANPR data of a real-world 39.25 km highway with four pairs of ANPR detectors of the uplink and downlink, Zhejiang, China. The experimental results indicate that LSTM-CNN learns spatial, temporal, and depth information better than the state-of-the-art traffic forecasting models, so LSTM-CNN can predict more accurate travel time. Moreover, LSTM-CNN outperforms the state-of-the-art methods in nonrecurrent prediction, multistep-ahead prediction, and long-term prediction. LSTM-CNN is a promising model with scalability and portability for highway traffic prediction and can be further extended to improve the performance of the advanced traffic management system (ATMS) and advanced traffic information system (ATIS).

Day-to-Day Traffic Assignment Model considering Information Fusion and Dynamic Route Adjustment Ratio

A new day-to-day traffic assignment model is proposed to describe travelers’ day-to-day behavioral changes with advanced traffic information system. In the model, travelers’ perception is updated by a double exponential-smoothing learning process combining experience and traffic information that is explicitly modelled. Route adjustment ratio is dynamically determined by the difference between perceived and expected utilities. Through theoretical analyses, we investigate the existence of its fixed point and the influence factors of uniqueness of the fixed point. An iterative-based algorithm that can solve the fixed point is also given. Numerical experiments are then conducted to investigate effects of several main parameters on its convergence, which provides insights for traffic management. In addition, we compare the system efficiencies under the static route adjustment ratio and dynamic route adjustment ratio and show the application of the model.

Exploring the use of advanced traffic information system to manage traffic congestion in developing countries

Abstract Presently, advanced traveler information system (ATIS) is extensively used elsewhere and is regarded as an important component for road traffic control and management. ATIS uses information and communication technology to broadcast traveler information to commuters to help them in scheduling their trips and also offer directions on the way. Information provided by ATIS may include: traffic condition (which may be presented in queue length, travel time, or delay), incidents such as road traffic crash and their geographical locations, events such as road construction and demonstrations, optimal routes, and inclement weather which may disrupt traffic. The traffic information may be transmitted via mechanisms such as radio and television, dynamic navigation systems (in-vehicle or mounted), internet services, smart phone services (e.g. specially designed application) and variable speed limit or changeable message signs systems. Providing commuters with accurate, timely and reliable traffic information will influence travel behavior and ultimately result in reduced travel time and traffic congestion along with reduced vehicle emissions and fuel consumption. Notwithstanding the benefits which can be derived from using intelligent transport systems to manage traffic, its use can be described as being in the basic stage in most developing countries. This paper discusses how ATIS can play a vital role of efforts to reduce traffic congestion in emerging nations.

Forecasting the Short-Term Traffic Flow in the Intelligent Transportation System Based on an Inertia Nonhomogenous Discrete Gray Model

The traffic-flow system has basic dynamic characteristics. This feature provides a theoretical basis for constructing a reasonable and effective model for the traffic-flow system. The research on short-term traffic-flow forecasting is of wide interest. Its results can be applied directly to advanced traffic information systems and traffic management, providing real-time and effective traffic information. According to the dynamic characteristics of traffic-flow data, this paper extends the mechanical properties, such as distance, acceleration, force combination, and decomposition, to the traffic-flow data vector. According to the mechanical properties of the data, this paper proposes four new models of structural parameters and component parameters, inertia nonhomogenous discrete gray models (referred to as INDGM), and analyzes the important properties of the model. This model examines the construction of the inertia nonhomogenous discrete gray model from the mechanical properties of the data, explaining the classic NDGM modeling mechanism in the meantime. Finally, this paper analyzes the traffic-flow data of Whitemud Drive in Canada and studies the relationship between the inertia model and the traffic-flow state according to the data analysis of the traffic-flow state. A simulation accuracy and prediction accuracy of up to 0.0248 and 0.0273, respectively, are obtained.

Route Choice Model Based on Game Theory for Commuters

The traffic behaviours of commuters may cause traffic congestion during peak hours. Advanced Traffic Information System can provide dynamic information to travellers. Due to the lack of timeliness and comprehensiveness, the provided information cannot satisfy the travellers’ needs. Since the assumptions of traditional route choice model based on Expected Utility Theory conflict with the actual situation, a route choice model based on Game Theory is proposed to provide reliable route choice to commuters in actual situation in this paper. The proposed model treats the alternative routes as game players and utilizes the precision of predicted information and familiarity of traffic condition to build a game. The optimal route can be generated considering Nash Equilibrium by solving the route choice game. Simulations and experimental analysis show that the proposed model can describe the commuters’ routine route choice decisionexactly and the provided route is reliable.

Analysis of inertial choice behavior based expected and experienced savings from a real-world route choice experiment

In the context of route choice, inertial behaviour shows that drivers make choices that are satisfactory rather than optimal. Consequently, drivers may not necessarily alter their choice when confronted with a travel time increase on the current choice or a travel time decrease of a choice alternative. As a consequence it can be assumed that driver only alter their choice when the utility difference between alternatives becomes larger than some individualsituation-specific threshold. Route choice data from a real-world driving experiment was used to study inertial behaviour and estimate inertia thresholds. The data analysis consisted of two parts. One addressing expected travel time savings relative to choice alternatives and another addressing experienced travel time savings on the current route. With regard to the former it was found that on average roughly one-fourth of the choices were inertial choices. Small travel time differences and dominant non-travel time route attributes had a positive effect on the frequency of inertial choices. Based on lost travel time, inertia thresholds up to 4.5 minutes or 30% of the average travel time were found, while thresholds up to 1.3 minutes or 13% were most common. Considerable differences between OD-pairs and routes indicated that thresholds are probabilistic and dependent on the choice set. Considering experienced savings, it was found that on average roughly half of the choices concerned inertial behaviour. Inertia thresholds up to 3.31 minutes or 37% of the average travel time were found, while thresholds up to 1.18 minutes or 11% were more common. Switching behaviour could not be explained by experienced savings as participants were much less responsive to experienced loss on the current route as opposed to expected loss relative to the choice alternative. Future research opportunities are: moving-average thresholds, inertia thresholds of different situations and driver behaviour types, asymmetry between gains and losses, impact of advanced traffic information systems, and estimation of a model that better matches the route choice data than a simple model that is based on travel time alone does.

Traffic flow forecasting based on a hybrid model

Real-time traffic flow forecasting is of great importance in the development of advanced traffic management systems and advanced traveler information systems. Traffic flow is evaluated using time series, and the Autoregressive Integrated Moving Average (ARIMA) model has been commonly used for determining the regression-type relationship between historical and future data. However, the performance of the ARIMA model is limited by the difficulty of capturing nonlinear patterns and the challenges of diagnosing permanent white noises. Hence, a hybrid method of ARIMA-EGARCH-M-GED was developed with the intent to address those limitations. It combines the linear ARIMA model with a nonlinear model of Exponent Generalized Autoregressive Conditional Heteroscedasticity (EGARCH) to capture heteroscedasticity (the variance of random error varying across the data) of traffic flow series. EGARCH in Mean (EGARCH-M), which corrects the expression of conditional variance by connecting the conditional mean directly with the variance, was introduced to better restrain the influence of abnormal data. Moreover, the tail of the generalized error distribution (GED) is better than that of the normal distribution in characterizing the features of time series, especially heteroscedasticity of residual sequences. Data collected from an interstate highway (I-80 in California) with a sampling period of 5 minutes were used to evaluate the performance of the proposed model. The results from the hybrid model were compared with ARIMA, an artificial neural network, and a K-nearest neighbor model. The results showed that the hybrid model outperformed the other methods in terms of accuracy and reliability. Overall, the proposed model performed well in tracking the features of measured data and controlling the impact of abnormal data.

Agent-based en-route diversion: Dynamic behavioral responses and network performance represented by Macroscopic Fundamental Diagrams

This paper focuses on modeling agents’ en-route diversion behavior under information provision. The behavior model is estimated based on naïve Bayes rules and re-calibrated using a Bayesian approach. Stated-preference driving simulator data is employed for model estimation. Bluetooth-based field data is employed for re-calibration. Then the behavior model is integrated with a simulation-based dynamic traffic assignment model. A traffic incident scenario along with variable message signs (VMS) is designed and analyzed under the context of a real-world large-scale transportation network to demonstrate the integrated model and the impact of drivers’ dynamic en-route diversion behavior on network performance. Macroscopic Fundamental Diagram (MFD) is employed as a measurement to represent traffic dynamics. This research has quantitatively evaluated the impact of information provision and en-route diversion in a VMS case study. It proposes and demonstrates an original, complete, behaviorally sound, and cost-effective modeling framework for potential analyses and evaluations related to Advanced Traffic Information System (ATIS) and real-time operational applications.

Time Pressure Impact on En-Route Choice Behavior under Guidance Information

This paper investigates time pressure impact on en-route choice behavior under guidance information. The impact of time pressure constraints is quite evident in en-route choice decisions. Understanding en-route choice behavior under time pressure and predicting route choices are important components in the overall goal of building a more reliable and efficient advanced traffic information systems (ATIS). A hybrid model to predict en-route driver routing decisions under guidance information is proposed. The model accounts for the observed changes in choice probabilities, including preference reversals as a function of time limit.

A support vector machine with the tabu search algorithm for freeway incident detection

Abstract Automated Incident Detection (AID) is an important part of Advanced Traffic Management and Information Systems (ATMISs). An automated incident detection system can effectively provide information on an incident, which can help initiate the required measure to reduce the influence of the incident. To accurately detect incidents in expressways, a Support Vector Machine (SVM) is used in this paper. Since the selection of optimal parameters for the SVM can improve prediction accuracy, the tabu search algorithm is employed to optimize the SVM parameters. The proposed model is evaluated with data for two freeways in China. The results show that the tabu search algorithm can effectively provide better parameter values for the SVM, and SVM models outperform Artificial Neural Networks (ANNs) in freeway incident detection.

Modelling road users’ behavioural change over time in stochastic road networks with guidance information

This paper proposes a new dynamic model to describe road users’ day-to-day behavioural changes with guidance from a Reliable Path Searching System (RPSS) in the Advanced Traffic Information Systems. When given a desired on-time arrival probability, the RPSS provides a suggested travel time budget (TTB) to road users for planning their trips in the stochastic road network with day-to-day traffic demand variation. A discrete-time day-to-day learning mechanism is adopted to capture how road users perceive the stochastic travel times from day-to-day. A Bayesian dynamic process is used to model the adjustments of road users’ TTBs with guidance from the RPSS. Both the learning mechanism and the Bayesian dynamic process are incorporated into the day-to-day dynamic model. The properties of the day-to-day dynamic model are then presented, and a solution algorithm is proposed to solve the day-to-day dynamics. Numerical examples are used to illustrate the applications of the day-to-day dynamic model and the proposed algorithm. The effects of the accuracy level of the guidance information on road users’ behavioural change are also investigated through sensitivity tests. In addition, the day-to-day dynamic model is further used to examine the day-to-day dynamics when there are changes in road capacities due to temporary roadwork or network expansion. This paper represents a conjecture on how road users adjust their TTBs and route choices with reference to the desired on-time arrival probabilities in stochastic road network. And the numerical results on road users’ behavioural change over time can provide insights into the evolution trajectories of traffic state over time for consistent evaluation of network performance.

Vehicle Detector Evaluation Based on Concepts of Traceability and Confidence Interval

Vehicle detectors are essential devices for advanced traffic management and information systems. Hence, the accuracy of traffic data from detectors is fundamentally important. When evaluating detectors, one uses a baseline data source that generates reference data. However, reference data also contain errors and uncertainties, albeit small ones. Nevertheless, little research has been performed to address the errors and uncertainties of reference data. This study proposed a new detector evaluation scheme to resolve that problem. The scheme takes errors and uncertainties of reference data into account for reliable detector evaluation. Moreover, the scheme applies the interval estimation technique, instead of the traditional point estimation, to present detector evaluation results. The suggested scheme was thoroughly verified by using the South Korea detector test bed, a verification that the scheme produced reliable detector evaluation results.

Estimation of carbon dioxide emissions per urban center link unit using data collected by the Advanced Traffic Information System in Daejeon, Korea

CO2 emissions on roads in urban centers substantially affect global warming. It is important to quantify CO2 emissions in terms of the link unit in order to reduce these emissions on the roads. Therefore, in this study, we utilized real-time traffic data and attempted to develop a methodology for estimating CO2 emissions per link unit. Because of the recent development of the vehicle-to-infrastructure (V2I) communication technology, data from probe vehicles (PVs) can be collected and speed per link unit can be calculated. Among the existing emission calculation methodologies, mesoscale modeling, which is a representative modeling measurement technique, requires speed and traffic data per link unit. As it is not feasible to install fixed detectors at every link for traffic data collection, in this study, we developed a model for traffic volume estimation by utilizing the number of PVs that can be additionally collected when the PV data are collected. Multiple linear regression and an artificial neural network (ANN) were used for estimating the traffic volume. The independent variables and input data for each model are the number of PVs, travel time index (TTI), the number of lanes, and time slots. The result from the traffic volume estimate model shows that the mean absolute percentage error (MAPE) of the ANN is 18.67%, thus proving that it is more effective. The ANN-based traffic volume estimation served as the basis for the calculation of emissions per link unit. The daily average emissions for Daejeon, where this study was based, were 2210.19 ton/day. By vehicle type, passenger cars accounted for 71.28% of the total emissions. By road, Gyeryongro emitted 125.48 ton/day, accounting for 5.68% of the total emission, the highest percentage of all roads. In terms of emissions per kilometer, Hanbatdaero had the highest emission volume, with 7.26 ton/day/km on average. This study proves that real-time traffic data allow an emissions estimate in terms of the link unit. Furthermore, an analysis of CO2 emissions can support traffic management to make decisions related to the reduction of carbon emissions.

A Kalman Filter Approach for Exploiting Bluetooth Traffic Data When Estimating Time-Dependent OD Matrices

Time-dependent origin–destination (OD) matrices are essential input for dynamic traffic models such as microscopic and mesoscopic traffic simulators. Dynamic traffic models also support real-time traffic management decisions, and they are traditionally used in the design and evaluation of advanced traffic traffic management and information systems (ATMS/ATIS). Time-dependent OD estimations are typically based either on Kalman filtering or on bilevel mathematical programming, which can be considered in most cases as ad hoc heuristics. The advent of the new information and communication technologies (ICT) provides new types of traffic data with higher quality and accuracy, which in turn allows new modeling hypotheses that lead to more computationally efficient algorithms. This article presents ad hoc, Kalman filtering procedures that explicitly exploit Bluetooth sensor traffic data, and it reports the numerical results from computational experiments performed at a network test site.

Corridor-level signalized intersection safety analysis in Shanghai, China using Bayesian hierarchical models

Most traffic crashes in Chinese cities occur at signalized intersections. Research on the intersection safety problem in China is still in its early stage. The recent development of an advanced traffic information system in Shanghai enables in-depth intersection safety analyses using road design, traffic operation, and crash data. In Shanghai, the road network density is relatively high and the distance between signalized intersections is small, averaging about 200m. Adjacent signalized intersections located along the same corridor share similar traffic flows, and signals are usually coordinated. Therefore, when studying intersection safety in Shanghai, it is essential to account for intersection correlations within corridors. In this study, data for 195 signalized intersections along 22 corridors in the urban areas of Shanghai were collected. Mean speeds and speed variances of corridors were acquired from taxis equipped with Global Positioning Systems (GPS). Bayesian hierarchical models were applied to identify crash risk factors at both the intersection and the corridor levels. Results showed that intersections along corridors with lower mean speeds were associated with fewer crashes than those with higher speeds, and those intersections along two-way roads, under elevated roads, and in close proximity to each other, tended to have higher crash frequencies.

Weighing communication overhead against travel time reduction in advanced traffic information systems

Advanced traffic information systems can assist drivers in reducing their travel times by making better use of available road capacity. In assessing their practical applicability, however, it is important to assess the overhead that various advanced traffic information systems bring. This paper evaluates the communication overhead for a decentralized, delegate multi-agent system based advanced traffic information system and for a centralized system. We document the relationship between the communication overhead and travel time reduction for both systems. This analysis can help in weighing both factors when designing a practical traffic information system in a real-world scenario.

Exploring Link Covering and Node Covering Formulations of Detection Layout Problem

The primary data input used in principal traffic models comes from origin–destination (O-D) trip matrices, which describe patterns of traffic behavior across a network. O-D matrices are a critical requirement in advanced traffic management or information systems that are supported by dynamic traffic assignment models. However, because O-D matrices are not directly observable, current practice adjusts an initial seed matrix from link flow counts that are provided by an existing layout of traffic-counting stations. The adequacy of the detection layout is critical to determining the quality of the adjusted O-D matrix. The usual approaches to the detection layout problem assume that detectors are located at network links. This paper proposes a modified set that formulates the link detection layout problem with side constraints and presents a new metaheuristic tabu search algorithm with high computational efficiency. Emerging information and communication technologies (ICT), especially those based on the detection of the electronic signature of onboard devices (such as Bluetooth devices), allow the location of sensors at intersections. To take into account explicitly how these ICT sensors operate, this paper proposes a new formulation of a node-covering problem with side constraints, which for practical purposes can be efficiently solved with standard professional solvers such as CPLEX.