2013 Session: 736

• Modeling Emission Policies Through Travel Demand Mechanisms: Analysis of the Best Reduction Strategies
  Abstract: Emission reduction strategies are gaining attention as planning agencies work towards adherence to air quality conformity standards. As state departments of transportation (DOTs) and metropolitan planning organizations (MPOs) struggle to find more options to reduce Greenhouse Gases (GHG), emission pricing offers a solution. To consider emission pricing as an alternative, planners and policymakers will need tools to understand the implications on travel behavior of private vehicle users. In this paper we present an integrated travel demand and emission model to incorporate policy strategies for emission reduction. First, the travel demand model determines the destination, mode and route choice of the users in response to a particular strategy set by the planner. Second, the emission model provides GHG (NOx, VOC, and CO2) estimates at a very detailed level in the transportation network. A logit-based destination choice and mode choice is proposed and the user’s response to a strategy in assignment is captured by Frank-Wolfe algorithm. A Base-case and four models are proposed to achieve emission reduction in a multimodal transportation network. Each model provides several insights to pollutant specific emission, how different function classes of the network are affected by policies, impacts on vehicle miles of travel (VMT), total system emission, and total system travel time. The complete framework is applied to Montgomery County’s (located in the Washington DC-Baltimore region in the United States) multimodal transportation network. It is observed that each model has a set of advantages and limitations in terms of emission reduction.
  Authors: Welch, Timothy F.; Mishra, Sabyasachee
  Authors: Welch, Timothy F.; Mishra, Sabyasachee
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0148
• Road Congestion in Europe: Measurement and Monitoring Using GPS Traces
  Abstract: The methodology presented here allows to measure and monitor road congestion across Europe using traces from in-vehicle navigation systems. The approach is based on the analysis of a large number of real vehicle speeds that have been measured on each road link and the application of algorithms that allow the estimation of congestion indicators for specific types of roads during selected time periods. The results include the detailed mapping of recurrent congestion both geographically and temporally, as well as the comparison of the quality of service of road networks between different zones.The large number of "probes" (over 1 trillion of measurements) gives a highly accurate and representative picture of the actual driving conditions across the European road network. The data collected are clustered in groups of speed profiles which represent change in average speed behaviour along a road link in five-minute time intervals over a 24-hour period. Each road link has a specific speed profile assigned per day of the week. The average speed on a specific link during a certain time period can be compared to the benchmark speed estimated for the link under free flow conditions or against selected threshold values. As a result, indicators of congestion for different time periods can be measured and compared across links, regions and countries. The paper also presents examples of the application of this methodology:• Mapping congestion and monitoring its evolution over time, by comparing the level and distribution of congestion in two different points in time • Application of the congestion indicators in European transport policy, by comparing average congestion between the peak hour and wider time periods and identifying measures to improve the temporal distribution• Combination of congestion indicators with traffic counts in order to improve speed flow curves used in transport network models
  Authors: Christidis, Panayotis; Ibañez Rivas, Nicolás
  Authors: Christidis, Panayotis; Ibañez Rivas, Nicolás
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0679
• Mixed Multinomial Logit Model for Out-of-Home Leisure Activity Choice
  Abstract: This paper documents the design and results of a study on the factors influencing the choice of out-of-home leisure activities. Influencing factors seem related to socio-demographic characteristics, personal preferences, characteristics of the built environment and other aspects of the activities themselves such as transport mode and day of the week. The Dutch 2008-2009 Continuous Time Use Research data (CVTO) on leisure activities is used to estimate a mixed multinomial logit model that accounts for heterogeneity in individuals’ preferences. The present model formulation also allows the analysis of substitution and complementary between the different types of activities. Results indicate differences in behavior between people with different socio-demographic characteristics, especially regarding their lifecycle stages, which is a variable composed by age and household composition. The built environment exerts a small influence on leisure activity purpose choice, whereas travel party has a strong influence. In terms of substitution and complementary relationships between activity purposes it was found that while outdoor leisure decreases the propensity to perform sports or hobbies/courses activities, it increases the propensity to perform fun shopping and going out/culture activities. A discussion of limitations of the study completes the paper
  Authors: Grigolon, Anna Beatriz; Kemperman, Astrid; Timmermans, Harry J.P.
  Authors: Grigolon, Anna Beatriz; Kemperman, Astrid; Timmermans, Harry J.P.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0683
• Automated Inference of Linked Transit Journeys in London Using Fare-Transaction and Vehicle-Location Data
  Abstract: Urban public transit providers have historically planned and managed their networks and services with limited knowledge of their customers’ travel patterns. While ticket gates and bus fareboxes yield counts of passenger activity in specific stations or vehicles, the relationships between these transactions—the origins, transfers, and destinations of individual passengers—have typically been acquired only through costly and infrequent rider surveys. Building upon recent work on the utilization of automated fare-collection and vehicle-location systems for passenger-behavior analysis, this paper presents methods for inferring the journeys of all riders on a large public transit network.Using complete daily sets of data from London’s Oyster farecard and iBus vehicle-location system, boarding and alighting times and locations are inferred for individual bus passengers, interchanges (transfers) are inferred between passenger trips of various public modes, and origin–destination matrices of linked intermodal transit journeys are constructed, which include the estimated flows of non-farecard passengers. The outputs are validated against surveys and traditional origin–destination matrices, and the software implementation demonstrates that the procedure is efficient enough to be performed daily, enabling transit providers to observe travel behavior on all services at all times.
  Authors: Gordon, Jason B.; Koutsopoulos, Haris N.; Wilson, Nigel H.M.; Attanucci, John
  Authors: Gordon, Jason B.; Koutsopoulos, Haris N.; Wilson, Nigel H.M.; Attanucci, John
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0740
• Investigation of Required Travel Survey Size for Training an Activity-Based Traffic Demand Model for Flanders Implemented in the FEATHERS Simulation Platform
  Abstract: It is known from many years now that operational activity-based models need a lot of survey data to incorporate behavioural decision making of people. While there have been contributions from the field of statistics about how much survey data is needed to come to reliable estimates of behaviour; an obvious question which is often overlooked in the domain is how much survey data is really necessary to obtain an activity-based model that is sufficiently competent and accurate. This question is not only scientifically challenging and interesting, but also can significantly reduce data collection costs and is also very useful for practitioners. A very appealing question would be whether an activity-based model could also be trained with a smaller survey data set without losing too much model quality. This paper tries to explore this research question in the case of an activity-based model for Flanders (Belgium) inside the ‘Forecasting Evolutionary Activity-Travel of Households and their Environmental RepercussionS’ (FEATHERS) framework. As the scheduler in this study is based on decision trees, progressive sampling is being applied in order to investigate accuracy for all discrete choice decision trees. Based on the results of this investigation, it is demonstrated that for some decision trees the activity-based survey data set can be very small without losing accuracy, while for other decision trees bigger data sets are needed.
  Authors: Kochan, Bruno; Bellemans, Tom; Janssens, Davy; Wets, Geert
  Authors: Kochan, Bruno; Bellemans, Tom; Janssens, Davy; Wets, Geert
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0864
• Estimating Service-Related Traffic Demand from Trip Chain Data
  Abstract: Commercial traffic constitutes a significant part of total traffic. While on long distances goods traffic prevails, particularly in metropolitan areas service related traffic (i.e. traffic resulting from services delivered to customers at home, offices, constructions sites etc.) takes the lead – and is continuously growing.It is common practice to forecast transport demand for passenger traffic and goods traffic. Dedicated models for service related traffic are still rare. This is not only due to the recognition of the particularities of this special part of traffic among transport planners, it is also due to the complexity of analyzing and depicting service related traffic.One characteristic of service related traffic is that it consists of tours connecting sometimes more than one destination with the origin of the tour. Furthermore, many trips are trips with commercial vehicles, registered by a company or organization, but also private vehicle owners use their cars to some extent for commercial trips. This poses the opportunity to use trip chain data available for private and commercial vehicles to derive total transport demand of service related trips from a sample population.This article introduces such a methodology to extrapolate traffic demand from trip chain data. It is not only applicable to service related traffic, but to all tour based traffic, if empirical tour patterns are available. A regular survey of the German Federal Motor Transport Authority together with detailed spatial and demographic data was used as input to the model. The city of Berlin served as a case study.
  Authors: Schneider, Sebastian; Wolfermann, Axel J.
  Authors: Schneider, Sebastian; Wolfermann, Axel J.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-1181
• Predicting Tour Patterns Derived from Ubiquitous Data Sources
  Abstract: Ubiquitous data sources (e.g. GPS traces, mobile phone call records, smart card entries etc.) are increasingly being popular among researchers for deriving mobility patterns. These mobility patterns depend on socio-economic characteristics of the traveler as well as associated situational, contextual and environmental factors. These patterns therefore vary substantially with age, gender, employment status, income level and other demographic factors. Habitual markers (propensity to use the internet, social networking websites, mobile phones or media for example) also provide useful indications about these patterns. In this research, the day-to-day tour patterns of travelers have been extracted from GPS and WLAN records and a discrete choice modeling framework has been proposed to predict these patterns using demographic factors, habitual markers and other travel related attributes collected from travelers of Lausanne, Switzerland. The model parameters are estimated by maximum likelihood technique using the software BIOGEME. Estimated model parameters confirm that demographic factors (e.g. age, occupation and gender) combined with habitual markers (e.g. habit of listening to music on smart phone) and other contextual attributes (e.g. day of the week) can be used to predict the tour pattern of an individual on a certain day. The developed model demonstrates how information obtained from ubiquitous data sources can be successfully used as a tool for transportation planning and management.
  Authors: Iqbal, Shahadat; Siddique, Abu Bakkar; Islam, Md Mozahidul; Choudhury, Charisma Farheen
  Authors: Iqbal, Shahadat; Siddique, Abu Bakkar; Islam, Md Mozahidul; Choudhury, Charisma Farheen
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-1330
• Optimization Method of Alternate Traffic Restriction Scheme Based on Elastic Demand and Mode Choice Behavior
  Abstract: To cope with urban traffic congestion problem, alternate traffic restriction (ATR) has become an emerging countermeasure, in which a certain proportion of automobiles are prohibited to enter the pre-determined ATR districts in specific time periods. This study proposes an optimization method of ATR scheme for both its restriction districts and restriction proportion of automobiles. As a Stackelberg game between traffic policy makers and road users, the ATR scheme optimization problem is established with a bi-level programming model. The upper-level model is about ATR scheme aiming at consumer surplus maximization under the condition of overload flow minimization. The lower-level model is to optimize the elastic demands, mode choice and multi-class user equilibrium assignment synthetically. The proposed genetic algorithm with prolonging codes is of high computing efficiency in that it dynamically includes newly-appeared overload links into the codes so as to reduce the following searching range. Moreover, practical processing approaches are suggested to improve the operability of the model-based solutions. To our knowledge, this study is the first attempt to theoretically optimize the ATR scheme by systematic approach with mathematical model specification.
  Authors: Shi, Feng; Xu, Guang-ming; Liu, Bing; Huang, Helai
  Authors: Shi, Feng; Xu, Guang-ming; Liu, Bing; Huang, Helai
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-1853
• Adjusting Origin-Destination Matrices Based on Traffic Counts and Bootstrapping Confidence Intervals
  Abstract: Mobility studies require, as a preliminary step in most of them, conducting a survey to a sample of users of the transportation system. The statistical reliability of the data determines the goodness of the results and conclusions which can be inferred from the analyses and models generated. Due to the high economic costs of the collecting field stages, collected data are partially reused in either a disaggregated o aggregated manner. In the first case, the statistical reliability is not always guaranteed, affecting drastically the results to be derived from projections and estimates of future and hypothetic scenarios.In this paper we present a methodology, based on the techniques of "bootstrap", for the robust statistical estimation of the mobility matrices, and generate the confidence intervals of travel between origin-destination (OD) pairs defined by each matrix cell derived from a mobility survey. This result is of interest in defining the dimensions of certinty for matrix cells and subsequent adjustment by techniques based on aggregate data (i.e. traffic counts, cordon line matrices, paths, etc..).Bootstrap techniques was invented in the 70's and, although widely used since the 90's, have not been professionally full exploited in the field of mobility studies matrices.To address this task we have counted with a statistically reliable data mobility study conducted in Spain at the level of regions. This paper presents the results derived from disaggregating date at interprovincial level, and an application to the posterior mobility matrix adjusment based on traffic counts data. The study results demonstrate the potential of the methodology developed and the usefulness of conclusions.
  Authors: Benitez, Francisco Garcia; Romero, Luis M.; Caceres, Noelia; Del Castillo, Jose Maria
  Authors: Benitez, Francisco Garcia; Romero, Luis M.; Caceres, Noelia; Del Castillo, Jose Maria
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-1947
• Social Networks in Agent-Based Models for Carpooling
  Abstract: In this paper we present social networks in an agent-based model (ABM) for carpooling. Our model for the carpooling application is a computational model for simulating the interactions of autonomous agents and for analysing the effects of change in factors related to the infrastructure, behaviour and cost. Primarily, we focus on our agent-based approach for creating social networks for the carpooling application using socio-demographic data and daily activity-trip schedules estimated by Feathers, which is an activity-based traffic demand model. Social networks for the carpooling application, called carpooling SocNet in this paper, depicts the potential relationship information between carpoolers. We need relationship data to initiate our agent communication model and then employ a route matching algorithm and a utility function to trigger the negotiation process between agents. To generate carpooling SocNet, we proposed three similarity measures: profile, path and time interval similarity measure. In order to test the three similarity measures, we conducted experiments with input data in the Hasselt region and Limburg province, Belgium. As a result, it shows an interesting relationship information between the agents, which people in the study area have 65% of similarity to each other based on socio-economic attributes. Moreover, we found it is important to find an optimal value of the threshold because of the impact on finding a carpool partner and dependency on the study area. We plan to, as a part of the future work, use this carpooling SocNet data and feed it to our agent-based model to initiate communication, coordination and negotiation in carpooling.
  Authors: Cho, Sungjin; Yasar, Ansar-Ul-Haque; Knapen, Luk; Patil, Bharat; Bellemans, Tom; Janssens, Davy; Wets, Geert
  Authors: Cho, Sungjin; Yasar, Ansar-Ul-Haque; Knapen, Luk; Patil, Bharat; Bellemans, Tom; Janssens, Davy; Wets, Geert
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2055
• Do Others Influence Our Travel Decision?
  Abstract: Based on Japanese panel data derived from 12 monthly waves and 1,253 questionnaires, the study examines the impact of social interactions on tourism participation. Social interactions are classified into three types: namely, endogenous, exogenous and correlated effects. It is empirically confirmed that endogenous social effects have significant influences on tourism participation behavior. For example, it is found that interactions with people of the same income groups (an example of endogenous social effects) are generally significant across most months of the year. However, it is worth noting that endogenous social effects will be overestimated if the correlated social effect is not taken into account. The results suggest that social multiplier effects can be discerned that have potential implications for promotion and management of tourism participation rates.
  Authors: Wu, Lingling; Zhang, Junyi
  Authors: Wu, Lingling; Zhang, Junyi
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2282
• Paired Route Impedance Correction for Multinomial Logit Model Based on Equivalent Impedance
  Abstract: A closed-form logit-style formulation basing on route impedance correction is proposed to alleviate defeats caused by the independence of irrelevant alternatives (IIA) and the homoscedasticity properties of multinomial logit (MNL) route choice model. The algorithm utilizes the traditional correction method to add addition impedance to each route by route pair combination with improved correction algorithm. For each route pair, a binary logit model with the concept of Logit Equivalent Impedance is utilized to calculate the route impedance correction, which is derived from two assumptions: (1) the route choice probabilities are independent of the overlapping part so that IIA property can be alleviated; and (2) the variances of route impedances are proportional to route impedances to resolve the homoscedasticity issue. The closed-form structure and easy computation of original MNL model remain unchanged. Numerical examples show that the proposed approach produces more reasonable results than traditional models with same complexity of computation, and is more stable when the number of routes in reasonable route choice set changes.
  Authors: Jun, LI; Lai, Xinjun; Yu, Zhi
  Authors: Jun, LI; Lai, Xinjun; Yu, Zhi
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2486
• Evaluating Impact of Congestion Pricing on Greenhouse Gas Emission Reductions: Study of Beijing
  Abstract: The rapidly growing number of motorized vehicles in Beijing puts a high pressure on the city’s transport systems. At the same time, greenhouse gas emissions and local air pollutants as well as external costs of traffic congestion are increasing. City government and decision makers are increasingly interested in understanding the effectiveness of travel demand management measures to reduce congestions and considering as an important instrument for meeting the carbon emission reduction targets. This paper described and demonstrated a detailed bottom-up approach for evaluating greenhouse gas emission impacts along a congestion pricing measure. The approach chosen is to link greenhouse gas emission evaluation to travel activity modeling. A state-of-the-practice regional travel demand model was used to simulate the travel effects of proposed congestion pricing scenarios on the vehicle kilometers traveled (VKT) by various modes (vehicle type) and road network Level of Service performance. Combined with energy intensity based emission factors, the emissions are computed as a function of road type, vehicle type, fuel type, traffic volume and network traffic situations. The congestion charging measure was estimated to reduce traffic related carbon dioxide emissions by 19.4 percent and fuel consumption by 21.2 percent within the charging zone, and in the city proper by 2.4 percent. Congestion pricing program for Beijing is technically feasible to ease traffic congestion and reduce greenhouse gas emissions.
  Authors: Sun, Shengyang; Li, Chunyan; Bongardt, Daniel; Wen, Huimin
  Authors: Sun, Shengyang; Li, Chunyan; Bongardt, Daniel; Wen, Huimin
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2693
• Day-to-Day Origin-Destination Tuple Estimation and Predictionwith Hierarchical Bayesian Networks Using Multiple Data Sources
  Abstract: Predicting traffic demand becomes essential, either to understand what the traffic state is in the future or to take necessary measures for alleviating the congestion in the next time period. Usually, an origin destination matrix (OD) is used to represent traffic demand between two zones in transportation planning. Vehicles are assumed to be homogenous and the trips of each vehicle are examined separately. In fact, this traditional OD-matrix lacks of a behavioral basis and trip based model structure. There is additionally another research stream of travel activity-based research which digs into the individual travel behaviors. The stream really takes care of the trip chain for travelers. But their research scope is on the attributes of the trips, ignoring the road network. In order to link these two fields and to better predict traffic demand, we propose the concept of Origin Destination Tuple (ODT), as a sequence dependent OD pairs. Although the ODT requires additional assumptions to the prediction process, the help of advanced monitoring systems to identify and track vehicles in the road network can mitigate additional uncertainties, reducing the under-specification more specifically. We propose a Hierarchical Bayesian Networks mechanism in the Gaussian Space to get the posterior of uncertain parameters. The model includes level and trend components to make prediction of future traffic volumes. A case study demonstrates that the proposed method is feasible to predict the demand and the path flow from cameras can reduce the uncertainty in the estimation and prediction process, especially for the OD-tuple.
  Authors: Ma, Yinyi
  Authors: Ma, Yinyi
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2861
• Do Your Neighbors Affect Your Mode Choice? Spatial Probit Model for Commuting to Ohio State University
  Abstract: Neighborhood effects have recently become a focus of interest in transportation research, whereby transportation mode choice is not only affected by an individual’s characteristics and the physical conditions of the transportation system, but also by the mode choices of that individual’s neighbors. This study supports the neighborhood effects argument, using a spatial econometrics approach and data from The Ohio State University’s 2011 Campus Transportation Survey. A spatial probit model of commuters’ mode choices (auto versus non-auto) is estimated, accounting for spatial autocorrelation. The results reveal that the more non-auto (walking, bicycling, and transit) users are residing around an individual, the more attractive these modes become for this individual. In addition to these spatial effects, the results indicate that students are more likely to commute to campus by non-auto modes, as compared to faculty and staff, and that the probability of choosing non-auto modes decreases with distance from campus. Feeling of safety, duration of travel, flexibility of departure time, ability to make stops on the way to/from campus, and attitudes towards auto use (being a car patron or a captive user), also affect an individual’s mode choice. These findings provide campus transportation planners new insights on the factors influencing travel mode choices.
  Authors: Wang, Chih-Hao; Akar, Gulsah; Guldmann, Jean-Michel
  Authors: Wang, Chih-Hao; Akar, Gulsah; Guldmann, Jean-Michel
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2907
• Convergence Analysis of Markov Chain Monte Carlo Estimators of a Transportation Mode Choice Model
  Abstract: Although Bayes estimators are attractive for problems involving non-convex optimization and for weakly identified discrete choice models, researchers in travel behavior modeling seem somewhat reluctant to adopt the Bayesian framework. A common argument against simulation-based Bayes estimators is that there are no general rules for assessing convergence. In this paper, we study convergence of the Markov chain Monte Carlo (MCMC) estimator of a discrete choice model via the empirical analysis of travel mode choice using a series of commonly used diagnostics. We also compare the behavior of the posterior first and second moments with the point estimates of maximum simulated likelihood. Some of the tests are more conservative than others, suggesting the use of a longer chain (500,000 iterations for the specific empirical problem being analyzed), while other test provide evidence in favor of a shorter chain (50,000 iterations). The comparison with the maximum simulated likelihood estimator shows that the only problem with 50,000 iterations is the loss of efficiency in the determination of the nuisance parameters.
  Authors: Daziano, Ricardo A.; Wang, Chen
  Authors: Daziano, Ricardo A.; Wang, Chen
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3233
• Travel Time Modeling with Taxi GPS and Household Survey Data
  Abstract: Transportation planning agencies all over the world usually maintain travel time matrices (or network skims) for a small number of time windows. In order to predict travelers' response to congestion mitigation strategies, it becomes essential to develop time of day choice models that require travel time estimates at a finer time resolution. In this paper, we develop regression models to relate travel times collected from taxi GPS data to the network travel times and compare the results to a similar model estimated with household travel survey data. The rationale behind this procedure is to develop a formula that allows the calculation of travel time for any origin-destination pair and for any time of the day, given the network travel times for three time periods (AM peak, PM peak, and off-peak). The two data sources, survey and GPS data, are compared based on descriptive statistics and by plotting the variation of the predicted speed by time of day. Statistical tests are performed to investigate whether the two data sources can be pooled together. The test results indicate that though there are significant differences in the estimated coefficients which do not vary across time of day (for example, coefficients of distance, and central business district indicators), the two data sources exhibit comparable profiles of time of day variation in speed up to a certain scale.
  Authors: Li, Siyu; Enam, Annesha; Abou Zeid, Maya; Ben-Akiva, Moshe E.
  Authors: Li, Siyu; Enam, Annesha; Abou Zeid, Maya; Ben-Akiva, Moshe E.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3297
• Route Choice Modelling for Urban Commuters: Considering Bridge Choice as a Key Determinant of Selected Routes
  Abstract: Trip assignment is still a modelling and prediction challenge. For aggregate analyses, traditional trip assignment approaches may suffice. However, investigations of drivers’ choices with respect to network infrastructure changes require more disaggregate and behavioural approach. Effects of critical infrastructure elements in the network on route choice behaviour of the drivers are often crucial to investigate. The case of Montreal is of particular interest since the city, an Island, is completely separated from the rest of the region by two important rivers. Consequently, drivers have to select one of the available bridges to reach their destination. The research relies on a set of observed trips with bridge declaration from a large-scale travel survey conducted in 2008. It is a one-day trip diary reaching some 4% of the residing population and including the bridge chosen in the itinerary for car driver trips. The paper provides a descriptive analysis of the bridges and their usage. An advanced discrete choice model that jointly models choice set formation and final choice is then formulated and estimated using the observed trips. Empirical model correctly identifies effects of travel time interacting with time of day and destination trip purpose. Travellers are more sensitive to travel time during off-peak period. Empirical results show that age, gender and household auto ownership explain the variation of scale parameters of route/bridge choice; for instance, older people show more stable route/bridge choice behaviour than younger ones. Discussion on the performance of the model is provided along with further result analysis and perspectives for further work.
  Authors: Nurul Habib, Khandker M.; Morency, Catherine; Trepanier, Martin; Salem, Sarah
  Authors: Nurul Habib, Khandker M.; Morency, Catherine; Trepanier, Martin; Salem, Sarah
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3534
• How Captive Is the Captive Market Anyway? Reexamination of the Impact of Auto Availability
  Abstract: The concept of the “captive market” for transit has been prevalent in transportation planning agencies for decades. Indeed, many transit agencies focus considerable effort on distinguishing between their “choice” and “captive” markets. This paper does not intend to undermine the theory that a segment of the market, predominantly lower-income riders, is more constrained in their travel choices. However, the paper does argue that the concept of the “captive market” can be applied in an overly deterministic way when members of 0 auto households are essentially locked out of certain travel modes such as auto driver (and/or single occupancy vehicle mode) and drive to transit (park-and-ride). Data from two recent household trip diaries (2008 and 2011) conducted in metropolitan Vancouver are examined to determine whether such rule-based approaches are appropriate for 0-car households, particularly in light of the rise of auto-sharing companies in Vancouver and surrounding cities. The paper will provide an analysis of two different travel patterns of particular relevance for individuals that fall broadly into the captive market category: the car-availability for travelers not using auto modes will be examined in addition to the mode choice of travelers from 0-car households. A detailed examination will be made of those respondents from 0-car household who also indicated that they were a car driver.
  Authors: Petersen, Eric
  Authors: Petersen, Eric
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3550
• Can Transport Models Predict Effects of Congestion Charges? Ex-post Evidence on Forecast Accuracy in Stockholm
  Abstract: This paper compares forecasted effects of the Stockholm congestion charges with actual outcomes. We show that the transport model was able to predict the charges’ effects on travel demand reasonably well, but with some underprediction of the effect on leisure trips. The most important concerns during the design of the congestion charging scheme were traffic reduction in bottlenecks, increase in public transport ridership, decrease of vehicle kilometres in the city centre, and traffic effects on circumferential roads. All of these factors were predicted well enough to allow planners to draw correct conclusions regarding the design and preparations for the scheme. The one major shortcoming was that the static assignment model was unable to predict the substantial effects on travel times. We conclude that the transport model worked well enough to be useful as decision support, performing considerably better than (unaided) “experts’ judgment”, but that results must be interpreted taking the model’s various limitations into account. The positive experiences from the Stockholm congestion charges hence seem to be transferable to other cities in the sense that if a charging system is forecasted to have beneficial effects on congestion, then this is most likely true.
  Authors: Eliasson, Jonas; Börjesson, Maria Magdalena; van Amelsfort, Dirk Hendrik; Brundell-Freij, Karin; Engelson, Leonid
  Authors: Eliasson, Jonas; Börjesson, Maria Magdalena; van Amelsfort, Dirk Hendrik; Brundell-Freij, Karin; Engelson, Leonid
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3608
• A Multi-Model ApproachtTo Large-Scale Multiagent Transport Simulation
  Abstract: In this paper, we introduce a multi-model approach to a large-scale, activity-based, multi-agent travel demand simulation.The Multi-Agent Transport Simulation toolkit, MATSim, is a full activity-based travel demand model, capable of handling very large urban scenarios in the order of millions of commuters. Its greatest current performance limitation is the network loading simulation, currently a queue simulation (`QSim'). In our application, the multi-model system periodically replaces the current QSim for a number of iterations with a simplified pseudo-simulation (`PSim') that runs approximately two orders of magnitude faster. PSim uses information generated in the preceding QSim iteration to produce an estimate of how well an agent day plan might perform, which allows the existing model framework to select and improve plans before executing them in a full queue simulation. We test the technique in an extensive scenario for Zurich, Switzerland, incorporating mode choice, road-pricing, secondary activity location choice, activity timing adjustment and dynamic routing. We find that the technique dramatically improves convergence rates for such complex, large-scale simulations, and fully exploits modern multi-core computer architectures. Its simple operational logic promises easy integration with all existing and upcoming MATSim functionality, and opens the door to more sophisticated approaches to large-scale, integrated transportation planning.
  Authors: Fourie, Pieter J.; Illenberger, Johannes; Nagel, Kai
  Authors: Fourie, Pieter J.; Illenberger, Johannes; Nagel, Kai
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3631
• Reliable Short-Term Traffic Flow Forecasting for Urban Roads Using Multivariate GARCH Model
  Abstract: Short-term traffic flow forecasting plays an essential role to the advanced traveler information systems, route guidance systems, and proactive traffic signal control systems. Therefore, numerous univariate and multivariate models have been presented on either traffic flow level forecasting or traffic flow variance forecasting. However, few studies have incorporated the relationship between different traffic parameters (such as volume and speed) into the traffic flow forecasting model development. Based on the well-known facts that there are some inherent relationships between traffic parameters and that there exists the heteroscedasticity in traffic flow series, a vector autoregressive (VAR) plus multiple generalized autoregressive conditional heteroscedasticity (MGARCH) method was proposed in this paper for reliable short-term traffic flow forecasting for urban roads, in which the VAR model was used as the mean equation of the MGARCH model for modeling the traffic flow levels and the MGARCH model was used to model the conditional traffic flow variances. The proposed method was validated and evaluated using the actual traffic volume and speed data. Evaluation results showed that the proposed method can generate worktable performance in terms of forecasting accuracy and the forecasted confidence intervals.
  Authors: Xia, Jingxin; Nie, Qinghui; Huang, Wei; Qian, Zhendong
  Authors: Xia, Jingxin; Nie, Qinghui; Huang, Wei; Qian, Zhendong
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3634
• Analyzing Sources of Variability in Travel Time Use in a Combined Framework Using Extended Structural Equation Models and UK National Travel Survey Data
  Abstract: This paper presents a new approach to investigating comprehensively the influences upon the variations in personal travel times. The influences include socio-economic and demographic characteristics of the travellers and their households, accessibility, land use characteristics of their residential areas, and the interactions between different trip purposes. Existing literature has variously examined such influences, but so far as we are aware this is the first time that all the above influences are examined within one combined framework. This is made possible through the use of extended structural equation models. The work presented here is focused on home to work journeys and shopping which takes the bulk of the personal travel times. The methodology is a general one and the investigation can be further expanded to include other travel purposes in the investigation, and indeed to investigate personal travel distances or trip rates. We use the UK National Travel Survey data from 2003 to 2008 which is a consistent and sufficiently large sample for our purposes. At the national level, we find that over the six years from 2003 to 2008 there is no evidence of any significant trend in travel time spent, either for home to work journeys or shopping, provided that appropriate explanatory variables are included in the model. At a personal level, we find that there is a wide range of variables that exert significant influences both directly and indirectly. Because socio-economic and demographic variables have a pronounced effect on residential location patterns and because all of these in turn influence car ownership rates, there is naturally a high degree of inter-correlation between the set of such influences on travel time, which is why the SEM-based approach has proved essential. The findings also show significant interactions among the trip purposes which should be considered for inclusion within advanced travel demand models.
  Authors: Jahanshahi, Kaveh; Jin, Ying; Williams, Ian N.
  Authors: Jahanshahi, Kaveh; Jin, Ying; Williams, Ian N.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4287
• The Modeling of Household Vehicle Type Choice Accommodating Spatial Dependence Effects
  Abstract: Household vehicle ownership and fleet composition are choice dimensions that have important implications for policy making, particularly in the energy and environmental sustainability arena. In the context of household vehicle ownership and type choice, it is conceivable that there are substantial spatial interaction effects due to both observed and unobserved factors. This paper presents a multinomial probit model formulation that incorporates spatial spillover effects arising from both observed and unobserved factors. The model is capable of endogenously estimating the number of vehicles of each type that a household acquires by using a synthetic choice occasion approach where households are assumed to purchase vehicles over a series of choice occasions. The model is estimated on the California add-on data set of the 2009 National Household Travel Survey. Model estimation results show that spatial dependency effects are statistically significant. The findings have important implications for model development and application in the policy forecasting arena.
  Authors: Paleti, Rajesh; Bhat, Chandra R.; Pendyala, Ram M.; Goulias, Konstadinos G.
  Authors: Paleti, Rajesh; Bhat, Chandra R.; Pendyala, Ram M.; Goulias, Konstadinos G.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3920
• Ordered and Unordered Discrete-Continuous Models: Comparative Analysis for Household Vehicle Holding and Mileage Traveled Decisions
  Abstract: Integrated models for household vehicle holding and mileage travelled are used for strategic planning and are necessary to analyze policies concerning congestion management, land use, fuel consumption, energy pricing, and pollution. A number of studies have demonstrated that unordered behavioral models perform better than ordered mechanisms for vehicle holding decisions. Those comparative studies have been conducted for the discrete part only and are often of logit type. Probit type models are usually adopted for joint discrete-continuous decisions for the flexibility offered by the multivariate normal to capture correlations across the two independent variables. Ordered probit models are in general preferred to unordered probit for the saving in computational costs deriving from the closed mathematical form of the choice probabilities. In this study, we extend to ordered probit a previous discrete-continuous model based on a density estimation approach with unrestricted correlation between the discrete and the continuous parts. A comparative analysis is then performed using data extracted from the 2001 and 2009 NHTS data. Estimation results show that discrete-continuous ordered probit are superior to unordered structures in terms of goodness of fit. Model applications for policy analysis reveals that density and driving cost only affect marginally vehicle holding decisions and annual miles driven. Those results seem to be stable across 2001 and 2009.
  Authors: Cirillo, Cinzia; Liu, Yangwen; Tremblay, Jean-Michel
  Authors: Cirillo, Cinzia; Liu, Yangwen; Tremblay, Jean-Michel
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3965
• Calibration of BPR Function Based on Link Counts and Its Sensitivity to Varying Demand
  Abstract: Travel demand models provide critical input while making investment decisions for large-scale transportation projects. Therefore, enhancing the accuracy of these models has been an important research area. As used in this paper, accuracy is the ability to reproduce road volume as close as possible to the observed road counts. Accuracy can be increased by improving any of the four steps (trip generation, distribution, mode split, and traffic assignment) used to develop transportation travel demand models (TDM). This paper is focused on the volume delay functions (VDFs) that are needed to implement the last step, traffic assignment. Once a travel demand model is acceptably validated with its base condition to the observed road counts, modelers use the same setting for VDFs to predict congested travel times based on future volumes which in most cases increase due to population growth. This paper utilizes a previously developed genetic algorithm to find the best VDF parameters for a given demand level and then uses those parameters for varying demands and compares the results with associated synthetic vehicle counts. Results showed that VDFs calibrated to higher demand generally perform better in varying demands than those calibrated in lower demand scenarios. The results of this study can help transportation planners in selecting VDFs that produce more accurate results.
  Authors: Foytik, Peter; Cetin, Mecit; Robinson, Robert Michael
  Authors: Foytik, Peter; Cetin, Mecit; Robinson, Robert Michael
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4019
• Reflecting Impacts of Systemwide Pricing Strategies in an Integrated Continuous-Time Prism-Constrained Activity-Travel Simulator of Demand and Supply
  Abstract: Pricing policies are increasingly being considered in urban areas around the world to better manage travel demand, mitigate adverse impacts of automobile travel, and raise financial resources for much needed infrastructure development. This paper demonstrates the feasibility of applying an integrated microsimulation model of activity-travel demand and dynamic traffic assignment for analyzing the impact of pricing policies on the entire range of activity-travel choices. The model system is based on a dynamic integration framework wherein the activity-travel simulator and the dynamic traffic assignment model communicate with one another along the continuous time axis so that trips are routed and simulated on the network as and when they are generated. This framework is applied to the analysis of a systemwide pricing policy for a small case study site to demonstrate how the model responds to various levels of pricing. A generalized travel time measure that accounts for the pricing effect and reflects population heterogeneity in values of time is used to simulate activity-travel choices, including both activity generation and destination choice, in a prism-constrained activity-travel simulation paradigm. Case study results show that trip lengths, travel time expenditures, and vehicle miles of travel are impacted to a greater degree than activity-trip rates and activity durations as a result of pricing policies. Measures of change output by the model are found to be consistent with elasticity estimates reported in the literature, suggesting that the model is able to reflect adjustments in activity-travel behavior in response to pricing policies.
  Authors: Konduri, Karthik Charan; Pendyala, Ram M.; You, Daehyun; Chiu, Yi-Chang; Hickman, Mark D.; Noh, Hyunsoo; Waddell, Paul; Wang, Liming; Gardner, Brian
  Authors: Konduri, Karthik Charan; Pendyala, Ram M.; You, Daehyun; Chiu, Yi-Chang; Hickman, Mark D.; Noh, Hyunsoo; Waddell, Paul; Wang, Liming; Gardner, Brian
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4025
• Geolocating Activities to Business Establishment Locations Using Time-Dependent Activity Assignment for Travel Demand Modeling
  Abstract: Activity geolocation is the identification of the real-world geographic location of an activity. It is closely related to destination choice and stop location choice in activity-based approaches in travel demand forecasting models. In this paper, a new activity assignment approach is proposed that considers as input the activities predicted for each person in an activity based microsimulation model system called SimAGENT and an inventory of business establishments provided by a commercially available database. This approach produces geolocated activities (e.g., eating out at a restaurant, going shopping) at locations of a subset of real-world business establishments enabling small area studies and the micro-assignment of greenhouse gas assignment at fine resolution to perform impact analysis. This method is implemented in TRansportation ANalysis SIMulation System (TRANSIMS) and compared to a naive approach of assigning activities to random locations within traffic analysis zones of SimAGENT.
  Authors: Tang, Daimin; Ravulaparthy, Srinath; Goulias, Konstadinos G.
  Authors: Tang, Daimin; Ravulaparthy, Srinath; Goulias, Konstadinos G.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4455
• Impact of Distribution Choice for Representing Input Variation in Travel Demand Uncertainty Analysis Simulation in the Context of Information Shortage
  Abstract: Uncertainty and risk analysis is becoming an ever increasing part of the transport demand forecasting since it can significantly influence the feasibility of a transportation project. Probabilistic assessment using Monte Carlo simulation is one of the most common approaches to uncertainty evaluation. The Monte Carlo method implies generating random draws from probability distributions for the input variables. Often the empirical data which allows fitting the distribution is not available at all or involves additional costs to be obtained. In this case the modeller assumes the probability distribution shape and its parameters in a context of information shortage and this introduces additional error to the uncertainty analysis. The main goal of this study is quantifying the impact of the distribution choice and, more specifically, of its shape, skewness and correlations among the variables, on the estimates of the model uncertainty present in the model attributes, applied to a study case of the High-Speed Railway project in Portugal. The results suggest that the mode location of the distribution, its shape and the correlations affect significantly the uncertainty analysis outcome.
  Authors: Petrik, Olga; de Abreu e Silva, João; Moura, Filipe
  Authors: Petrik, Olga; de Abreu e Silva, João; Moura, Filipe
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4688
• Modeling the Choice of Household Vehicle for Social-Recreational Tours
  Abstract: Predicting the vehicle used for individual trips and tours can help improve the quantification of energy consumption, quality of emissions forecasts, and assess impacts of policies that vary by vehicle type. The focus of this study is to contribute to incorporating this aspect within current activity-based modeling frameworks. Data from the 2009 National Household Travel Survey are used in this analysis. The empirical scope of this study is limited to two-adult two car households in the context of social-recreational tours; however, this methodology can be directly extended to other cases as well. An exploratory analysis indicates that the “primary driver” variable is perhaps the strongest predictor of the vehicle allocated to independent tours and tours made by adults with children. In the case of joint tours, there is clearly a choice of vehicle to be made. Following the exploratory analysis, two models were developed. One allocates each vehicle to a primary driver in the household (long-term, household-level model). The second allocates a vehicle for the joint tours (short-term, tour-level model). Both models were estimated using the unlabeled binary-logit methodology. Several vehicle attributes (such as size/body type, fuel efficiency, age, and operating costs) and socio-economic variables (age, and presence of children) were estimated to be statistically-significant predicators of the vehicle-allocation patterns.
  Authors: Lim, Kwangkyun; Srinivasan, Sivaramakrishnan
  Authors: Lim, Kwangkyun; Srinivasan, Sivaramakrishnan
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4789
• Toll Pricing: Computational Tests on How to Capture Heterogeneity of User Preferences
  Abstract: Motivated by the increasing interest in implementing and exploring a wider range of pricing alternatives and technologies, sometimes in conjunction with operational schemes, this paper compares different approaches for capturing the heterogeneity of user preferences in forecasting the demand for tolled facilities resulting from user responses to pricing schemes. Existing tools used in practice typically deal with users’ value of time (VOT) heterogeneity in one of two principal ways: ignore it, and use a constant value of time, or recognize it by defining discrete user classes that each correspond to a given VOT range, represented by its midpoint or average VOT. The first is fundamentally incorrect, and would lead to highly biased; the second, which is only a coarse approximation, is difficult to calibrate a priori, and may become unwieldy computationally. Recently, a third approach, which recognizes that VOT is a continuous variable that is distributed probabilistically across the user population, is gaining greater attention. We are interested in the following questions: (i) what are the impacts of different VOT assumptions on prediction biases in forecasts of toll road usage under different pricing schemes? (ii) do the discrete treatments provide a good approximation of models with continuously distributed VOT, and if so how should VOT be assigned to user classes? and (iii) what are the relative computational implications of different VOT assumptions, and what is the role in this regard of efficient implementation techniques for large-scale network applications? The results suggest that recognizing the continuous nature of the VOT distribution is justified on both accuracy and computational grounds, especially in light of recent algorithmic and implementation advances in this regard.
  Authors: Jiang, Lan; Mahmassani, Hani S.
  Authors: Jiang, Lan; Mahmassani, Hani S.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5035
• Deconstructing the D-variables: New Methods to Measure theBuilt Environment for Travel Behavior Research
  Abstract: Measures are determined by the data, technology and methods available. Until recently these have all limited travel behavior and built environment research to rely on zonally aggregated averages, homogeneously attributed to unique individuals. Furthermore, most studies focus on the characteristics of the origin, not the critical components of the destination (such as parking, availability, price, etc.), let alone the area in-between. But travel is an inherently linear activity with strong destination determinants, and thus these zonal measures of the trip origins likely miss key subtleties of the built environment important to people traveling outside of the protective enclosure of an automobile (bicyclists and pedestrians). This paper presents the development and statistical testing of new methods to more closely align detailed measures of the built environment with the individual¬– in sum, more finely disaggregated data of the built environment for disaggregated analyses of travel behavior. To ground the development of these new methods and measures, this paper applies their development to a real-world problem: How does the urban environment influence the probability that commuters will access rapid transit stations? Through the use of predictive, multinomial logit (MNL) mode choice models and MNL model comparison methods, this paper tests whether new, linear based built environment measures are an improvement over zonally aggregated “D-Variable” measures now commonly used in practice. These model comparison techniques prove the superiority of these measures over the D-Variables for linear spatial units of analysis.
  Authors: Appleyard, Bruce
  Authors: Appleyard, Bruce
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5146
• Sociodynamic Discrete Choice on Networks in Space: Role of Utility Parameters and Connectivity in Emergent Outcomes
  Abstract: The reported research treats social interactions and generated feedback dynamics in the adoption of various transportation mode alternatives. We consider a model where a commuter’s choice is directly influenced by the percentages of neighbors and socioeconomic peers making each choice, and which accounts for common unobserved attributes of the choice alternatives in the error structure. We explicitly address non-global interactions within different social and spatial network structures, combining econometric estimation with computational techniques from multi-agent based simulation, and present an empirical application of the model using pseudo-panel microdata collected by the Amsterdam Agency for Traffic, Transport and Infrastructure. The paper extends previous work by the authors in considering the effects of various hypothesized socio-geographic networks. We also test for the effect of the scale of the interaction, comparing municipal district clusters versus smaller 4-digit postcode clusters. We observe that the estimated utility parameters for the different sociogeographic network scenarios can generate dramatically different dynamics and thus cannot be ignored in any empirical application. However, in a hypothetical simulation experiment we find that swapping the sociogeographic networks does not significantly change the long-run outcome of the simulation, when utility parameters are held fixed. We conclude highlighting recommendations for future work.
  Authors: Dugundji, Elenna Rose; Gulyás, László
  Authors: Dugundji, Elenna Rose; Gulyás, László
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5214
• Introducing Shopping and Leisure Facilities: Study on Agent-Based Transport Modeling in South Africa
  Abstract: Agent-based transport modeling is a promising emerging technology that has already achieved success in a number of countries, including South Africa. Until recently South African research into agent-based modeling of traffic behavior mainly focused on commuters travelling between home and work. Although the models accurately predict travel times, people in the real world travel for a number of reasons other than work, such as going to school or going shopping. Agent-based models are often criticized as being data intensive, both in developing and developed countries, and consequently limited work has been done relating to secondary activities. In this paper we show how open data, in the form of OpenStreetMap, is used to enhance existing traffic models. We demonstrate how, in an agent-based setting, the model responds in a realistic way using shopping and leisure facility data in the Nelson Mandela Metropolitan in the Eastern Cape of South Africa. This is achieved by extending an existing agent-based model and providing agents with more realistic spread of behavioral responses, in this case a choice of facilities where secondary activities such as shopping and leisure can take place.
  Authors: Momsen, Sune; Joubert, Johan W.
  Authors: Momsen, Sune; Joubert, Johan W.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4621
• Agent-Based Approach for Integrating Departure Time and Dynamic Traffic Assignment Models
  Abstract: The dynamic analysis of transportation networks has received much attention through the past two decades. Likewise, departure time choice models have been well studied due to their importance in dealing with peak-period traffic congestion. This paper describes research on the integration of simulation-based DTA and a positive behavioral departure time model, which has been seldom discussed in the existing literature. The positive agent-based modeling approach in this paper requires longitudinal behavioral process data, but helps overcome certain limitations of traditional utility-based models and focuses more on how individuals actually make departure time and routing choices. The resulting agent-based modeling framework consists of three components: departure time choice, dynamic route choice, and dynamic network loading. By considering departure time and dynamic routing decisions together, we can more comprehensively evaluate traffic operations strategies and transportation planning scenarios. In the proposed agent-based model, fixed overall travel demand is assigned to departure time intervals and travel routes based on individuals’ day-to-day learning of the traffic pattern simulated with a DTA module. In a large-scale real-world demonstration, this agent-based approach is applied to analyze the peak spreading and route diversion effects of a new toll road project (with time-of-day pricing) in Maryland. The effect of time discretization, an important consideration in dynamic network/demand analysis, on model output is also explored and discussed.
  Authors: Zhang, Lei; Mollanejad, Mostafa; Xiong, Chenfeng; Zhu, Shanjiang
  Authors: Zhang, Lei; Mollanejad, Mostafa; Xiong, Chenfeng; Zhu, Shanjiang
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5218
• Equilibrium Traffic Assignment: New Model for Spatially Disaggregate Demand
  Abstract: In all traffic assignment procedures, Traffic Analysis Zones are represented punctually by their centroids. This is the cause for much error. In this paper, a new model for traffic assignment is introduced, that explicitly tackles this issue. This model does not necessitate changes in zones or centroid connectors. It relies on a careful modeling of connectors as random variables, calibrated using land-use databases. A mathematical description of the model is presented, along with a solution algorithm for equilibrium assignment. Finally, the model is applied to the Paris region, with encouraging results.
  Authors: Benezech, Vincent; Leurent, Fabien M.
  Authors: Benezech, Vincent; Leurent, Fabien M.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5321
• Paired Route Impedance Correction for Multinomial Logit Model Based on Equivalent Impedance
  Authors: Lai, Xinjun
  Authors: Lai, Xinjun
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2486
• Travel Time Modeling with Taxi GPS and Household Survey Data
  Authors: Li, Siyu
  Authors: Li, Siyu
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3297
• Introducing Shopping and Leisure Facilities: Study on Agent-Based Transport Modeling in South Africa
  Authors: Joubert, Johan
  Authors: Joubert, Johan
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4621
• Impact of Distribution Choice for Representing Input Variation in Travel Demand Uncertainty Analysis Simulation in the Context of Information Shortage
  Authors: Petrik, Olga
  Authors: Petrik, Olga
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-4688
• Sociodynamic Discrete Choice on Networks in Space: Role of Utility Parameters and Connectivity in Emergent Outcomes
  Authors: Dugundji, Elenna
  Authors: Dugundji, Elenna
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5214
• Modeling Emission Policies Through Travel Demand Mechanisms: Analysis of the Best Reduction Strategies
  Authors: Welch, Timothy
  Authors: Welch, Timothy
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0148
• Mixed Multinomial Logit Model for Out-of-Home Leisure Activity Choice
  Authors: Grigolon, Anna
  Authors: Grigolon, Anna
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0683
• Automated Inference of Linked Transit Journeys in London Using Fare-Transaction and Vehicle-Location Data
  Authors: Gordon, Jason
  Authors: Gordon, Jason
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0740
• Investigation of Required Travel Survey Size for Training an Activity-Based Traffic Demand Model for Flanders Implemented in the FEATHERS Simulation Platform
  Authors: Bellemans, Tom
  Authors: Bellemans, Tom
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-0864
• Adjusting Origin-Destination Matrices Based on Traffic Counts and Bootstrapping Confidence Intervals
  Authors: Benitez, Francisco
  Authors: Benitez, Francisco
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-1947
• Reliable Short-Term Traffic Flow Forecasting for Urban Roads Using Multivariate GARCH Model
  Authors: Xia, Jingxin
  Authors: Xia, Jingxin
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3634
• Do Your Neighbors Affect Your Mode Choice? Spatial Probit Model for Commuting to Ohio State University
  Authors: Akar, Gulsah
  Authors: Akar, Gulsah
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2907
• Deconstructing the D-variables: New Methods to Measure the Built Environment for Travel Behavior Research
  Authors: Appleyard, Bruce
  Authors: Appleyard, Bruce
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5146
• Optimization Method of Alternate Traffic Restriction Scheme Based on Elastic Demand and Mode Choice Behavior
  Authors: Xu, Guang-ming
  Authors: Xu, Guang-ming
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-1853
• Toll Pricing: Computational Tests on How to Capture Heterogeneity of User Preferences
  Authors: Jiang, Lan
  Authors: Jiang, Lan
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-5035
• Social Networks in Agent-Based Models for Carpooling
  Authors: Cho, Sungjin
  Authors: Cho, Sungjin
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2055
• The Modeling of Household Vehicle Type Choice Accommodating Spatial Dependence Effects
  Authors: Paleti, Rajesh
  Authors: Paleti, Rajesh
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3920
• How Captive Is the Captive Market Anyway? Reexamination of the Impact of Auto Availability
  Authors: Petersen, Eric
  Authors: Petersen, Eric
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-3550
• Evaluating Impact of Congestion Pricing on Greenhouse Gas Emission Reductions: Study of Beijing
  Authors: Sun, Shengyang
  Authors: Sun, Shengyang
  Year: 2013
  Document Type: Presentation; Poster
  Subject: Planning and Forecasting
  Session: 736
  Paper Number: 13-2693