2013 Session: 658

• Using Time-Based Metrics to Compare Crash Risk Across Modes and Locations
  Abstract: The objective of this work is to identify better metrics of exposure when comparing traffic crash risk across modes or across locations. We propose that total time travelled should be used for road user exposure to crash risk. The idea behind this is that travel time reflects the differences in speeds across different modes and hence should be used as the basic exposure metric from which crash risk based on other metrics, such as travel distance, can easily be derived. We also propose that when comparing crash risk of different modes across different locations the time based mode share should be used as an explanatory variable. By using mode share we are generalizing the safety in numbers concept which focuses on absolute numbers. This work presents a discussion on why these two metrics were chosen and how they are different from the commonly used metrics. Quantitative evidence for the choice of time based metrics is also presented using travel survey data to compare crash risk across modes and locations.
  Authors: Guler, Sukran Ilgin; Grembek, Offer; Ragland, David R.
  Authors: Guler, Sukran Ilgin; Grembek, Offer; Ragland, David R.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-0522
• Multivariate Spatial Models of Excess Crash Frequency at Area Level: Case of Costa Rica
  Abstract: Recently, areal models of crash frequency have being used in the analysis of various area-wide factors affecting road crashes. On the other hand, disease mapping methods are commonly used in epidemiology to assess the relative risk of the population at different spatial units. A natural next step is to combine these two approaches to estimate the excess crash frequency at area level as a measure of absolute crash risk. Furthermore, multivariate spatial models of crash severity are explored in order to account for both frequency and severity of crashes and control for the spatial correlation frequently found in crash data. This paper aims to extent the concept of safety performance functions to be used in areal models of crash frequency. A multivariate spatial model is used for that purpose and compared to its univariate counterpart. Full Bayes hierarchical approach is used to estimate the models of crash frequency at canton level for Costa Rica. An intrinsic Multivariate Conditional Autoregressive model is used for modeling spatial random effects. The results show that the multivariate spatial model performs better than its univariate counterpart in terms of the penalized goodness-of-fit measure Deviance Information Criteria. Additionally, the effects of the spatial smoothing due to the multivariate spatial random effects are evident in the estimation of excess equivalent property damage only crashes.
  Authors: Aguero-Valverde, Jonathan
  Authors: Aguero-Valverde, Jonathan
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-1061
• Road Safety Forecasts in Five European Countries Using Structural Time-Series Models
  Abstract: Modeling road safety development is a complex task, which needs to consider both the quantifiable impact of specific parameters, as well as the underlying trends that cannot always be measured or observed. The objective of this research is to apply structural time series models for obtaining reliable medium- to long-term forecasts of road traffic fatality risk, using data from five countries with different characteristics from all over Europe (Cyprus, Greece, Hungary, Norway and Switzerland). Two structural time series models are considered: (i) the local linear trend model and the (ii) latent risk time-series model. Furthermore, a structured decision tree for the selection of the applicable model for each situation (developed within the DACOTA research project) is outlined. First, the fatality and exposure data that are used for the development of the models are presented and explored. Then, the modeling process is presented, including the model selection process, the introduction of intervention variables and the development of mobility scenarios. The forecasts using the developed models appear to be realistic and within acceptable confidence intervals. The proposed methodology is proved to be very efficient for handling different cases of data availability and quality, providing an appropriate alternative from the family of structural time series models in each country. A concluding section providing perspectives and directions for future research is finally presented.
  Authors: Antoniou, Constantinos; Papadimitriou, Eleonora; Yannis, George
  Authors: Antoniou, Constantinos; Papadimitriou, Eleonora; Yannis, George
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-1786
• Developing Cost Estimation Models for Road Rehabilitation and Reconstruction
  Abstract: The average unit costs of road works vary substantially between countries, and even between projects in the same country, due to a number of factors. In this paper an effort is made to develop prediction models for the unit costs of road works that could be applied for a wide range of conditions in different countries.A specialized dataset was used, which was generated under a World Bank study that included road works contracts from 14 countries in Europe and Central Asia (ECA). Two techniques were used for model development: multiple regression analysis and artificial neural networks. As the major problem found with the data set was missing or incomplete data, classification trees were used as an intermediate step to evaluate the correctness of the selected parameters.Three models were developed using regression analysis, two for the unit cost of asphalt concrete and one for the cost per km of rehabilitation and reconstruction works. The models include as independent variables the price of diesel fuel, country Gross National Income, World Governance Index, Transparency International Corruption Perception Index, percent of local bidders participating in the tender, and climate conditions. The analysis using classification trees confirmed the appropriateness of the variables selected in the regression analysis. The models developed using artificial neural networks were superior compared to the regression models, using mostly the same parameters.The resulting models could be particularly useful at the strategic level, for planning and optimization of works on road networks in ECA countries.
  Authors: Cirilovic, Jelena; Vajdic, Nevena; Mladenovic, Goran; Queiroz, Cesar
  Authors: Cirilovic, Jelena; Vajdic, Nevena; Mladenovic, Goran; Queiroz, Cesar
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2037
• Modeling Large-Truck Safety Using Logistic Regression Models
  Abstract: Statistics shows that crashes involving large trucks are generally more severe than those involving other vehicles due to the size, weight, and speed differential between trucks and other vehicles. Given the critical position of trucking in the process of economic recovery and growth, it is urgent to improve truck safety and mitigate any negative impacts to non-truck vehicles. Statistical models have been used universally to identify the contributing factors to crash severities and estimate injury probabilities. These different methodologies, albeit addressing different issues, may provide mixed results and the estimate accuracy may vary.The primary objective of this research is to investigate the effects of key determents to crash severities involving large trucks and to explore the relationship between them. The secondary objective is to provide insight on statistical applications by evaluating three logistic regression models: multinomial logistic (MNL), partial proportional odds (PPO), and mixed logistic (ML) models. The model results show that the majority of the coefficient estimates are consistent across the models studied. A few exceptions include young drivers and the use of safety constraints, which are not statistically significant in the ML model. The goodness-of-fit and model predictive power indicates that the PPO model produced the results that more closely resembled observations.
  Authors: Qin, Xiao; Wang, Kai; Cutler, Chase E.
  Authors: Qin, Xiao; Wang, Kai; Cutler, Chase E.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2067
• Hotspot Identification Under Limited Information: Combined Probabilistic and Fuzzy Cluster-Based Approach
  Abstract: Hot-Spot Identification (HSID) requires crash history information such as annual crash counts, their severities etc and details such as annual traffic exposure and geometric design details. The most recommended HSID method— Empirical Bayes utilizes at least crash history and traffic volume information to develop Safety Performance Function (SPF), which is used to compute expected number of crashes at a given site. However, in the absence of systematic data collection and maintenance, information about geometric design and traffic volume is not only difficult to obtain, but also demands significant resources. In such circumstances, only crash-count based (CCB) HSID techniques, such as Crash Frequency (CF) method, Fatal Crash Frequency (FCF) method and Equivalent Property Damage Only (EPDO) methods may only be adopted even with their known limitations. In this article, the authors suggested a new method of HSID, using disaggregate crash history information in crash severity model. Based on the probabilities of crash severities by the major contributing factors, expected numbers of severe and fatal crashes are calculated. These expected crash counts are used to classify locations into two fuzzy clusters— a) black-spots and b) white-spots using Fuzzy C-Means (FCM) algorithm. The identified hotspots are ranked based on their mean departure from core of the hotspot cluster. Site consistency, Method consistency and Total rank differences tests are used to compare the performance of the method with other CCB-HSID techniques. Results show the robustness of the proposed FCM method as it performs well in all consistency tests.
  Authors: Bandyopadhyaya, Ranja; Mitra, Sudeshna
  Authors: Bandyopadhyaya, Ranja; Mitra, Sudeshna
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2379
• Reliable Game Model for Network Violator Interception Problem
  Abstract: This paper focuses on planning interceptor locations in a general transportation network to maximize the expected benefits from catching violators mixing in public traveler flow. We use travel distance of violators before intercepted and innocent public traveler flow encountered by violators to define the expected benefit by setting interceptors along a specific route. Two reliability-related characteristics are also integrated into the planning model to make it more practical. One is each interceptor (maybe a sensor, a checkpoint or something else) have a failure probability. Different failure scenarios may incur different layout decision of interceptors and investigation of failure can lead to a more reliable solution. The other is there is a ¡°game¡± between interceptor planner and violators. We assume violators will adjust their route choices according to the interceptor layout decided by planner. Logit choice model is used to account for the route adjustment conducted by violators. Consequently, a non-linear non-convex binary integer programming model is presented. We develop a Simulated Annealing (SA) algorithm to solve it. A set of numerical experiments are conducted to illustrate the computational efficiency of the proposed algorithm. Further, we analyze the sensitivity of disruption probability of interceptors to optimal objective function values and discuss how to determine the values of parameters in violator route choice model.
  Authors: An, Shi; Cui, Jianxun; Wang, Jian
  Authors: An, Shi; Cui, Jianxun; Wang, Jian
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2431
• The Language of Driving: Advantages and Applications of Symbolic Data Reduction for Naturalistic Driving Data Analysis
  Abstract: Recent advances in onboard vehicle data recording devices have created an abundance of naturalistic driving data. The amount of data exceeds the resources available to analyze it, forcing researchers to focus on analyses of "critical events," which are identified using simple heuristics. This critical event analysis eliminates the context that can be critical in understanding driver behavior, reducing the generalizability of the analysis. This work introduces a method of naturalistic driving data analysis that will allow researchers to examine entire datasets by reducing them by over 90%. The method utilizes a symbolic data reduction algorithm, Symbolic Aggregate approXimation (SAX), which reduces time-series data to a string of letters. SAX can be applied to any continuous measurement and SAX output can be reintegrated into a dataset to preserve categorical information. This work explores the application of SAX to speed and acceleration data from a naturalistic driving dataset and demonstrates SAX's integration with other methods that can begin to tame the complexity of naturalistic data.
  Authors: McDonald, Anthony D.; Lee, John D.; Aksan, Nazan S.; Dawson, Jeffrey; Tippin, Jon; Rizzo, Matthew
  Authors: McDonald, Anthony D.; Lee, John D.; Aksan, Nazan S.; Dawson, Jeffrey; Tippin, Jon; Rizzo, Matthew
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2947
• Comparison of Sichel and Negative Binomial Models in Estimating Empirical Bayes Estimates
  Abstract: Traditionally, transportation safety analysts have used the empirical Bayes (EB) method to improve the estimate of the long-term mean of individual sites and to identify hotspots locations. The EB method combines two different sources of information: (1) the expected number of crashes estimated via crash prediction models, and (2) the observed number of crashes at individual sites. Crash prediction models have extensively been estimated using a negative binomial (NB) modeling framework due to the over-dispersion commonly found in crash data. Recent studies have shown that the Sichel (SI) distribution provides a promising avenue for developing crash prediction models. The objective of this study is to examine the application of the SI model in calculating EB estimates. To accomplish the objective of the study, the SI models with a fixed/varying dispersion term are developed using the crash data collected at 4-lane undivided rural highways in Texas. The important conclusions can be summarized as follows: (1) the selection of the crash prediction model (i.e., the SI or NB model) will affect the value of weight factor used for estimating the EB output; (2) the identification of hazardous sites, using the EB method, can be different when the SI model is used. Finally, a simulation study designed to examine which crash prediction model can better identify the hotspot is recommended as our future research.
  Authors: Zou, Yajie; Lord, Dominique; Zhang, Yunlong; Peng, Yichuan
  Authors: Zou, Yajie; Lord, Dominique; Zhang, Yunlong; Peng, Yichuan
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2938
• Full Bayes Methods for Road Safety Studies: Does Prior Specification Matter?
  Abstract: This paper investigates the effect of prior assumptions when applying Full Bayes (FB) methods in road safety analysis. The effect of prior choice is evaluated in the accuracy of model parameters, hotspot identification, goodness-of-fit, and treatment effectiveness index in before-after studies. Particular attention is devoted to conditions with lack of data referenced as the low-mean and small-sample problem. In this research, informative, semi-informative, and no-informative priors were determined based on past published studies. Using a simulation framework, various scenarios of sample size and crash occurrence mean are evaluated. Simulated data is generated based on two real databases of divided/undivided rural highway segments in New York and Texas. Diverse sample mean values were obtained considering different time periods (number of years) and classifying accidents in injury-fatal and total accidents. Among other results, it was found that under low-mean and small sample conditions, the outcomes can be significantly biased. However, the introduction of informative priors can still make feasible observational before-after studies when working with small number of observations from treatment and/or comparison sites. Informative priors can help provide more accurate estimates of the treatment effectiveness. Finally, in accordance with previous works, it was shown that the inverse dispersion parameter is significantly affected by prior specifications; nevertheless, regression parameters, goodness-of-fit, and hotspot identification are only slightly sensitive to prior choices.
  Authors: Miranda-Moreno, Luis Fernando; Heydari, Mohammad; Amador-Jimenez, Luis
  Authors: Miranda-Moreno, Luis Fernando; Heydari, Mohammad; Amador-Jimenez, Luis
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3042
• Analyzing Different Functional Forms of the Varying Weight Parameter for Finite Mixture of Negative Binomial Regression Models
  Abstract: Previously, the weight parameter of the finite mixture of regression models has been assumed to be invariant of the characteristics of the observations under study. Recently, it has been shown that the weight parameter of the finite mixture of negative binomial (NB) models can be dependent upon the attributes of the sites. Since the selection of the functional form for weight parameter has a significant impact on the classification results, there is a need to investigate how different functional forms affect the estimation of the varying weight parameter and whether there is a common functional form that can be properly used to model the weight parameter for different crash datasets. The primary objective of this research is to investigate the effect of different functional forms on estimation of the weight parameter as well as the group classification. To accomplish the study objectives, ten different functional forms for the varying weight parameter were estimated using three different multilane rural highway segment datasets: Texas undivided data, Texas divided data and Washington divided data. The results of this study confirm that the selection of the functional form for weight parameter will affect the classification results significantly. Among ten different functional forms, one functional form stands out for the three datasets. Therefore, when using the finite mixture of NB models with varying weight parameters to analyze the crash data, it is suggested that transportation safety analysts should include Model 5 (which models the classification as a function of the segment length raised to a power) along with other alternative functional forms for describing the weight parameter and select the most appropriate functional form based on not only the goodness-of-fit statistics, but also the classification results.
  Authors: Zou, Yajie; Zhang, Yunlong; Lord, Dominique; Peng, Yichuan
  Authors: Zou, Yajie; Zhang, Yunlong; Lord, Dominique; Peng, Yichuan
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3929
• Evaluating Alternate Discrete Choice Frameworks for Modeling Crash Injury Severity
  Abstract: This paper focuses on the relevance of alternate discrete choice frameworks for modeling driver injury severity. The study empirically compares the ordered response and unordered response models in the context of driver injury severity in traffic crashes. The alternative modeling approaches considered for the comparison exercise include: for the ordered response framework- ordered logit (OL), generalized ordered logit (GOL) and for the unordered response framework - multinomial logit (MNL), nested logit (NL) and ordered generalized extreme value logit (OGEV) model. A host of comparison metrics are computed to evaluate the performance of these alternative models. To our knowledge, the study provides a first of its kind comparison exercise of the performance of ordered and unordered response models for examining the impact of exogenous factors on the driver injury severity. The research also captures the effect of potential underreporting on alternative choice frameworks by artificially creating an underreported data sample from the driver injury severity sample.The empirical analysis is based on the 2010 General Estimates System (GES) data base. The comparison exercise clearly highlights the superiority of the GOL model on the estimation and the validation sample in terms of data fit compared to the OL and MNL models. The estimation with the artificial underreported sample consistently obtains the wrong elasticities and these errors are substantially reduced for both GOL and MNL models with the correction measures for the thresholds/constants of these models based on the true aggregate shares. The most striking finding is the fact that the MNL model does not perform any better in the underreporting context. In fact, the GOL elasticity effects of underreported estimates with corrections are closer to the true elasticity effects than that of the MNL model. Overall, the results of the empirical comparison provide credence to the belief that an ordered systems that allow for exogenous variable effects to vary across alternatives offer superior fit compared to unordered systems in modeling driver injury severity.
  Authors: Yasmin, Shamsunnahar; Eluru, Naveen
  Authors: Yasmin, Shamsunnahar; Eluru, Naveen
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4081
• Evaluating Short-Term Traffic Volume Forecasting Models Based on Multiple Data Sets and Data Diagnosis Measures
  Abstract: Although several short-term traffic volume forecasting methods have recently been developed, there is currently a lack of the studies which focus on how to choose the appropriate prediction method based on the statistical characteristics of the dataset. This paper first diagnoses the predictability of four different traffic volume datasets using various statistical measures including: (1) complexity analysis methods such as the delay time and embedding dimension method and the approximate entropy method; (2) nonlinearity analysis methods like the time reversibility of surrogate data; and (3) long range dependency analysis techniques like the Hurst Exponent. Following the diagnosis of the datasets, three short term traffic volume prediction models are applied: (1) Seasonal Autoregressive Integrated Moving Average (SARIMA); (2) k Nearest Neighbor (k-NN); and (3) Support Vector Regression (SVR). The results from the statistical data diagnosis methods are then correlated to the performance results of the three prediction methods on the four datasets in order to arrive at some conclusions regarding how to choose the appropriate prediction method. Among the conclusions of the study in that regard is that SVR is more suitable for nonlinear datasets, while SARIMA and k-NN are more appropriate for linear datasets. The data diagnosis results are also utilized to shed light on how to select the parameters of the different prediction models such as the length of the training data set for SARIMA and SVR, the average number of nearest neighbors for k-NN, and the input vector length for k-NN and SVR. Key Words: Short-term traffic volume prediction; time series analysis; Seasonal Autoregressive Integrated Moving Average (SARIMA); k Nearest Neighbor (k-NN); Support Vector Regression (SVR); Statistical methods
  Authors: Lin, Lei; Wang, Qian; Sadek, Adel W.
  Authors: Lin, Lei; Wang, Qian; Sadek, Adel W.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3691
• Spatial Generalized Ordered-Response Model to Examine Highway Crash Injury Severity
  Abstract: This paper proposes a flexible econometric structure for injury severity analysis at the level of individual crashes that recognizes the ordinal nature of injury severity categories, allows unobserved heterogeneity in the effects of contributing factors, as well as accommodates spatial dependencies in the injury severity levels experienced in crashes that occur close to one another in space. The modeling framework is applied to analyze the injury severity sustained in crashes occurring on highway road segments in Austin, Texas. The results from our analysis underscore the value of our proposed model to accurately estimate variable effects.
  Authors: Castro, Marisol; Paleti, Rajesh; Bhat, Chandra R.
  Authors: Castro, Marisol; Paleti, Rajesh; Bhat, Chandra R.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3810
• Simplified Two-Stage Choice Set Formation Models Incorporating Observed Choice Set Data
  Abstract: The implementation of a theoretically sound, two-stage discrete-choice modelling paradigm incorporating probabilistic choice sets is impractical when the number of alternatives is large, which is a typical case in most spatial choice contexts. In the context of residential location choice, Kaplan, Bekhor and Shiftan (2009, 2011, 2012) (KBS) developed a semi-compensatory choice model incorporating data of individuals searching for dwellings observed using a customised real estate agency website. This secondary data is used to compute the probability of considering a choice set that takes the form of an ordered probit model. In this paper, we illustrate that the simplicity of the KBS model arises because of an unrealistic assumption that individuals’ choice sets only contain alternatives that derive from their observed combination of thresholds. Relaxing this assumption, we introduce a new probabilistic choice set formation model that allows the power set to include all potential choice sets derived from variations in thresholds’ combinations. In addition to extending the KBS model, our proposed model asymptotically approaches the classical Manski model, if a suitable structure is used to categorize alternatives. In order to illustrate the biases inherent in the original KBS approach, we compare it with our proposed model and the MNL model using a Monte Carlo experiment. The results of this experiment show that the KBS model causes biases in predicted market share if individuals are free to choose from any potential choice sets derived from combinations of thresholds.
  Authors: Zolfaghari, Alireza; Sivakumar, Aruna; Polak, John W.
  Authors: Zolfaghari, Alireza; Sivakumar, Aruna; Polak, John W.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3870
• Multimodal Public Transport Demand: Cointegration Time-Series Approach
  Abstract: In this paper we investigate demand in a multimodal public transportation context. Demand is expressed as a function of operational and macroeconomic factors and is analyzed using a time-series cointegration and error correction approache. This allows for treating non stationary data and for determining short term and long term elasticities and at the same time estimating the speed of convergence from the short to the long term effects. As expected, short run elasticities appear lower than the long run ones, possibly because in the short run changes in explanatory factors are smaller and because behavior is governed by resistance to change. Fare and Income appear to have the greatest impact on public transport demand and also the greatest difference between short and long run elasticities.
  Authors: Milioti, Christina; Karlaftis, Matthew G.
  Authors: Milioti, Christina; Karlaftis, Matthew G.
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3910
• Collision Propensity Index for Unsignalized Intersections: Structural Equation Modeling Approach
  Abstract: The objective of this paper is to develop a quantitative collision propensity index (CPI) that captures the overall propensity of a given surrounding environment to cause accidents at un-signalized intersections. Using structural equation modeling, the index can be estimated from observed geometric, vehicular, driver-related, and traffic-related characteristics. Utilizing the California Department of Transportation's data repository, information on 4388 collisions occurring at 2709 different intersections was collected and processed. A statistically significant converging structural equation model was found reflecting the safety impact of different surrounding elements/dimensions on driving behavior: The CPI provides (a) a basis for quantifying the effects of the aforementioned characteristics on traffic safety and/or incident properties, (b) a basis for comparing the differences between the dimensions affecting collision propensity based on different exogenous measures’ classification schemes and (c) ranking the corresponding un-signalized intersections for improved safety performance. The framework and methodology used to develop this index has the potential to support safety policy analysis and decision making.
  Authors: Schorr, Justin; Hamdar, Samer Hani; Vassallo, Terasa
  Authors: Schorr, Justin; Hamdar, Samer Hani; Vassallo, Terasa
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3915
• Bayesian Approach to Real-Time Traffic State Estimation Using Particle Probability Hypothesis Density with Appropriate Clutter Intensity
  Abstract: Prediction of traffic flow variables such as traffic volume, travel speed or travel time for a short time horizon is of paramount importance in traffic control. Hence, data assimilation process in traffic modeling for estimation and prediction plays a key role. However, the increasing complexity, non-linearity and presence of various uncertainties (both in the measured data and models) are important factors affecting the traffic state prediction. To overcome this problem, new methodologies have to be investigated. In this aim, we propose in this paper the use of Probability Hypothesis Density (PHD). This methodology is intensively studied, developed and improved for the purposes of multiple object tracking and consists in the recursive state estimation of several targets by using the information coming from an observation process. However, some issues need to be studied, especially the clutter (false alarm) intensity. The goal of this paper is to expose the potential of the PHD filters for real-time traffic state estimation and the choice of an appropriate clutter intensity. This investigation is based on a Cell Transmission Model (CTM) coupled with the PHD filter. It brings a novel tool to the state estimation problem and allows one to estimate the densities in traffic networks. In this work, we compare this PHD filter with the particle filter (PF) which has been successfully applied in traffic control and conclude that the PHD filter can be seen as a relevant alternative that opens new research avenues.
  Authors: Canaud, Matthieu; Lyudmila, Mihaylova; El Faouzi, Nour-Eddin; Billot, Romain; Sau, Jacques
  Authors: Canaud, Matthieu; Lyudmila, Mihaylova; El Faouzi, Nour-Eddin; Billot, Romain; Sau, Jacques
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4401
• Even Perfect Regressions May Not Tell the Effect of Interventions
  Abstract: Suppose that there was a ‘well specified’ regression model, one in which the correct predictor variables were combined into the correct function, and that the unknown parameters were estimated using good and plentiful data. Can such a model be used to predict what change in the response variable is caused by a change in one of the predictor variables? Surprisingly the answer is: “No.” In this paper I identify the condition that often frustrates the causal interpretation of well specified regressions. I show how this very condition led astray several authors who used regressions to estimate the role of speed in accident generation
  Authors: Hauer, Ezra
  Authors: Hauer, Ezra
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4477
• Intelligent Evaluation of Transportation Management Policies for Metropolitan Areas
  Abstract: Traffic congestion, delay, accident, air and sound pollution are main downsides of living in metropolitans these days. Cities managers are trying to improve life standards in these cities while an important aspect is to reduce traffic and corresponding problems. In this research, expert knowledge is used to identify and organize effective criteria and rank policies for mitigating traffic problems in an overpopulated city. Tehran, the capital of Iran, the biggest city in the Middle East and the 16th dense city in the world is a good example which is studied in this research. To model the policy making problem, significant elements are identified and their weights are calculated and used to prioritize the candidate policies, i.e. bus network improvement, development of subway network, development of road infrastructures and development of bicycle network, using Analytic Hierarchy Process (AHP) based on the knowledge and judgment of a group of experts. Important criteria are classified into clusters of benefits, costs and opportunities in which totally 22 elements form the knowledge tree. Results show that user cost, congestion reduction, fuel consumption and safety approximately form 50 percent of total weight among all studied elements. According to the research conclusion, development of subway networks is the most efficient policy to reduce traffic problems in an overpopulated city as it increases benefits, reduces costs and improves traffic and transport opportunities. Sensitivity analysis shows that development of subway network always remains the superior policy for all relative importance values assigned to benefit, cost and opportunity criteria.
  Authors: Isaai, Mohammad Taghi; Najaf, Pooya; Lavasani, Seyed Mohammad; Nezamianpour Jahromi, Hossein
  Authors: Isaai, Mohammad Taghi; Najaf, Pooya; Lavasani, Seyed Mohammad; Nezamianpour Jahromi, Hossein
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4610
• Bayesian Inference of Traffic Volumes Based on Bluetooth Data
  Abstract: The study of the relationship between macroscopic traffic parameters, such as flow, speed and travel time, is essential to the understanding of the behaviour of freeway and arterial roads. However, the temporal dynamics of these parameters are difficult to model, especially for arterial roads, where the process of traffic change is driven by a variety of variables. The introduction of the Bluetooth technology into the transportation area has proven exceptionally useful for monitoring vehicular traffic, and Bluetooth data is now being used for travel time estimations and as a novel approach for studying traffic demand.In this work, we propose an approach based on Bayesian networks for analyzing and predicting the complex dynamics of flow or volume, based on travel time observations from Bluetooth sensors. The spatio-temporal relationship between volume and travel time is captured through a first-order transition model, and a univariate Gaussian sensor model. The two models are trained and tested on travel time and volume data, from an arterial link, collected over a period of six days. To reduce the computational costs of the inference tasks, volume is converted into a discrete variable. The discretization process is carried out through a Self-Organizing Map.Preliminary results show that a simple Bayesian network can effectively estimate and predict the complex temporal dynamics of arterial volumes from the travel time data. Not only is the model well suited to produce posterior distributions over single past, current and future states; but it also allows computing the estimations of joint distributions, over sequences of states. Furthermore, the Bayesian network can achieve excellent prediction, even when the stream of travel time observation is partially incomplete.
  Authors: Nantes, Alfredo; Billot, Romain; Miska, Marc Philipp; Chung, Edward
  Authors: Nantes, Alfredo; Billot, Romain; Miska, Marc Philipp; Chung, Edward
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4838
• Development of Statistically Based Methodology for Analyzing Safety Treatments at Isolated High-Speed Signalized Intersections
  Abstract: Crashes at isolated, rural intersections, particularly those involving vehicles traveling perpendicular to each other, are particularly bad because of high speeds involved. Many transportation agencies are interested in reducing the number of crashes at these types of intersections. There are many engineering treatments to improve the traffic safety at isolated, high-speed signalized intersections. Intuitively, it is critical to know which safety treatment may be the most effective for a given set of selection criteria for a particular intersection. Without a well-defined decision methodology it is almost impossible to decide which safety countermeasure or a set of countermeasures would the best option. Additionally because of the very large number of possible intersection configurations as well as the varying amount, distribution and type of traffic, it would be impossible to develop a set of guidelines that could be used for all signalized intersections. Therefore, it was undertaken to develop a methodology whereby common countermeasures could be modeled and analyzed before being implemented in the field. Because of the dynamic and stochastic nature of the problem it was decided to employ microsimulation tools, such as VISSIM, for analyzing the countermeasures. A calibrated and validated microsimulation model of signalized intersection was used to model two common safety countermeasures. The methodology was demonstrated on a test site located just outside of Lincoln, Nebraska. The model was calibrated to the distribution of observed speeds collected at the test site. It was shown that the methodology could be used for the preliminary analysis of the safety treatments.
  Authors: Wojtal, Remigiusz Marcin; Rilett, Laurence Russell
  Authors: Wojtal, Remigiusz Marcin; Rilett, Laurence Russell
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-5070
• Panel Mixed Ordered Probit Fractional Split Model: Modeling Vehicle Speed on Urban Roads in Montreal, Canada
  Abstract: Vehicle operating speed measured on roadways is a critical ingredient for a host of analysis in the transportation field including transportation safety, traffic flow, geometric design, vehicle emissions, and road user route decisions. The current research effort contributes to literature on examining vehicle speed on urban roads methodologically and substantively. In terms of methodology, we formulate a new econometric model framework for examining speed profiles. The proposed model is an ordered response formulation of a fractional split model. The ordered nature of the speed variable allows us to propose an ordered variant of the fractional split model. The proposed formulation allows us to model the proportion of vehicles traveling in each speed range for the entire segment of roadway. Further, we employ a mixed version of the fractional split model to account for the influence of site-specific unobserved effects. The paper contributes substantively by estimating the proposed model using a unique dataset from Montreal consisting of weekly speed data (collected in hourly intervals) for about 50 local roads and 70 arterial roads. We estimate separate models for local roads and arterial roads. The model estimation exercise considers a whole host of variables including geometric design attributes, roadway attributes, traffic characteristics and environmental factors. The model results highlight the role of various street characteristics including number of lanes, presence of parking, presence of sidewalks, vertical grade, and bicycle route on vehicle speed proportions. The results also highlight the presence of site-specific unobserved effects influencing the side walk variables as well as the standard deviation on the propensity constant. The parameters from the modeling exercise are validated using a hold-out sample not considered for model estimation. The results indicate that the proposed panel mixed ordered probit fractional split model offer promise for modeling such proportional ordinal variables.
  Authors: Eluru, Naveen; Chakour, Vincent; Chamberlain, Morgan; Miranda-Moreno, Luis Fernando
  Authors: Eluru, Naveen; Chakour, Vincent; Chamberlain, Morgan; Miranda-Moreno, Luis Fernando
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-5141
• Road Safety Forecasts in Five European Countries Using Structural Time-Series Models
  Authors: Antoniou, Constantinos
  Authors: Antoniou, Constantinos
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-1786
• Developing Cost Estimation Models for Road Rehabilitation and Reconstruction
  Authors: Mladenovic, Goran
  Authors: Mladenovic, Goran
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2037
• Full Bayes Methods for Road Safety Studies: Does Prior Specification Matter?
  Authors: Amador-Jimenez, Luis
  Authors: Amador-Jimenez, Luis
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3042
• Evaluating Short-Term Traffic Volume Forecasting Models Based on Multiple Data Sets and Data Diagnosis Measures
  Authors: Sadek, Adel
  Authors: Sadek, Adel
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3691
• Spatial Generalized Ordered-Response Model to Examine Highway Crash Injury Severity
  Authors: Castro, Marisol
  Authors: Castro, Marisol
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3810
• Multimodal Public Transport Demand: Cointegration Time-Series Approach
  Authors: Milioti, Christina
  Authors: Milioti, Christina
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3910
• Collision Propensity Index for Unsignalized Intersections: Structural Equation Modeling Approach
  Authors: Hamdar, Samer
  Authors: Hamdar, Samer
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-3915
• Evaluating Alternate Discrete Choice Frameworks for Modeling Crash Injury Severity
  Authors: Yasmin, Shamsunnahar
  Authors: Yasmin, Shamsunnahar
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4081
• Bayesian Inference of Traffic Volumes Based on Bluetooth Data
  Authors: Nantes, Alfredo
  Authors: Nantes, Alfredo
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4838
• Development of Statistically Based Methodology for Analyzing Safety Treatments at Isolated High-Speed Signalized Intersections
  Authors: Wojtal, Remigiusz
  Authors: Wojtal, Remigiusz
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-5070
• Panel Mixed Ordered Probit Fractional Split Model: Modeling Vehicle Speed on Urban Roads in Montreal, Canada
  Authors: Chakour, Vincent
  Authors: Chakour, Vincent
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-5141
• Using Time-Based Metrics to Compare Crash Risk Across Modes and Locations
  Authors: Guler, Sukran
  Authors: Guler, Sukran
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-0522
• Modeling Large-Truck Safety Using Logistic Regression Models
  Authors: Qin, Xiao
  Authors: Qin, Xiao
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2067
• Intelligent Evaluation of Transportation Management Policies for Metropolitan Areas
  Authors: Najaf, Pooya
  Authors: Najaf, Pooya
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-4610
• Evaluating Double Poisson Generalized Linear Model
  Abstract:

  The objectives of this study are to: 1) examine the applicability of the double Poisson (DP) generalized linear model (GLM) for analyzing motor vehicle crash data characterized by over- and under-dispersion and 2) compare the performance of the DP GLM with the COM-Poisson GLM in terms of goodness-of-fit and theoretical soundness. The DP distribution has seldom been investigated and applied since its first introduction two decades ago. The hurdle of applying the DP is related to its normalizing constant (or multiplicative constant) which is not available in closed form. This study proposed a new method to approximate the normalizing constant of the DP with high accuracy and reliability. The DP GLM and COM-Poisson GLM were developed using two observed over-dispersed datasets and one simulated under-dispersed dataset. The performances of the NB GLM (for over-dispersion) and Poisson GLM (for under-dispersion) were also provided as reference. The modeling results indicate that the DP GLM with its normalizing constant approximated by the new method can handle crash data characterized by over- and under-dispersion. Its performance is comparable to the COM-Poisson GLM in terms of GOF, although COM-Poisson GLM provides a slightly better fit. For the over-dispersed data, the DP GLM performs similar to the NB GLM. This study also shows that the traditional Poisson GLM overestimates the standard errors of the coefficients when the data are characterized by under-dispersion. Considering the fact that the DP GLM can be easily estimated and computationally inexpensive, it offers a flexible and efficient alternative for researchers to model the count data.

  Authors: Zou, Yaotian; Geedipally, Srinivas Reddy; Lord, Dominique
  Authors: Zou, Yaotian; Geedipally, Srinivas Reddy; Lord, Dominique
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Operations and Traffic Management; Safety and Human Factors
  Session: 658
  Paper Number: 13-2138