2013 Session: 688

• An Alpha-Reliable Combined Mean Travel Time Measure for Modeling Routing Risk Behaviors in Stochastic Traffic Network
  Abstract: Based on the reliability budget and percentile travel time (PTT), this study proposes a new travel time measure named the combined mean travel time (CMTT) to model the travellers¡¯ routing risk-attitudes under the stochastic traffic network. CMTT here is defined as the convex combination of the conditional expectations of PTT-below and PTT-excess travel times. The former is designed as a risk-optimistic travel time measure, and the latter is a risk-pessimistic one. The central idea of CMTT is comprehensively illustrated and the differences among the existing travel time measures are analysed. The Wardropian combined mean traffic equilibrium (CMTE) model is formulated as a variational inequality and solved via an alternating direction algorithm nesting the extra-gradient projection process. Some quantitative and qualitative properties of CMTT and CMTE model are rigorously proved. In the end, numerical example based on an artificial network is performed to characterize the CMTE network.
  Authors: Zhang, Wenyi; Guan, Wei; Song, Liying; Sun, Huijun
  Authors: Zhang, Wenyi; Guan, Wei; Song, Liying; Sun, Huijun
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-0611
• Latent Class Choice Model of Heterogeneous Drivers' Route Choice Behavior Based on Real-World Experiment
  Abstract: The research presented in this paper develops a hierarchical two-level heterogeneous route choice model using real-world experimental data. The study addresses two limitations in route choice literature, namely: experiment reality and driver heterogeneity. Specifically, aside from random error components, almost all route choice models used in transportation engineering practice assume that drivers are homogeneous in the way they make their route choices and in the way they respond to information. Although this paper only studies the way drivers make route choices, the proposed framework is capable of incorporating the heterogeneity of driver responses to information. The models developed in this paper are based on a sample of 20 drivers who collectively made more than 2,000 real-world route choices. In the proposed model, the first level model uses driver demographic and personality traits together with the characteristics of the choice situation to predict a behavior type. Within the context of this paper, a behavior type connotes a metaphoric measure of driver aggressiveness in route switching behavior, and captures driver behavior heterogeneity. The second level of the model uses the predicted behavior type and the travel experiences of the driver to predict the driver’s route choice. The results of the developed models indicate that in general: 1) behavior types can be predicted from driver demographics, personality traits, and choice situation characteristics, 2) the predicted behavior types are significant in route choice models, and 3) route choice models based on the proposed framework demonstrate better fits than state-of-the-art general models.
  Authors: Tawfik, Aly M.; Rakha, Hesham
  Authors: Tawfik, Aly M.; Rakha, Hesham
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-1367
• Modeling Framework to Analyze Effect of Multiple Traffic Information Service Providers on Traffic Network Performance
  Abstract: Along with the increasing popularity of mobile smart devices such as smartphones and tablet PCs, traffic information is coming closer to drivers through many intelligent traffic apps. Many countries have multiple traffic information service providers (ISP) who make every effort to improve their service quality in order to allure more people to subscribe to their services. As such a commercial environment is developing now, a study on the effect of multiple ISPs on the road network performance is necessary. Hence a modeling framework of day-to-day dynamics in which multiple information service providers (ISPs) compete/cooperate with each other to enhance their subscriber service quality is developed in this study. A realistic information acquisition and learning mechanism that ensures the consistent updating of individually perceived day-by-day travel times is incorporated for driver behavior in this framework. A boundedly-rational behavior model was adopted for route choice decisions. The framework is capable of investigating the effects of any potential competition/cooperation of multiple ISPs in the traffic information market in terms of their information sharing strategies. Numerical experiments on a real network were conducted to analyze the impact of such interactions on the network performance. The results showed that a cooperative system is not necessarily the best for network performance and that there exists an optimal level of market penetration of traffic information services in transportation networks beyond which the benefits will no longer increase, or can even worsen.
  Authors: Yang, Inchul; Jayakrishnan, R.
  Authors: Yang, Inchul; Jayakrishnan, R.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-2392
• Modeling Combined Travel Choices of Electric Vehicle Drivers with a Variational Inequality Network Formulation
  Abstract: Government incentives and policies can make plug-in electric vehicles more attractive. However, determining the incentives which provide the greatest social benefits requires tractable modeling techniques that can incorporate the unique features of electric vehicles. This paper develops models describing the differences in travel behavior between electric and gasoline vehicle drivers, investigating both the temporal and spatial travel choices behavior of these drivers. A multi-class quasi-dynamic model formulation is presented in the form of variational inequality. An optimization-based heuristic method is adopted to solve the model. Finally, numerical experiments are conducted to show that the model generates a network equilibrium solution of the combined choices.
  Authors: Zhang, Ti; Boyles, Stephen; Waller, S. Travis
  Authors: Zhang, Ti; Boyles, Stephen; Waller, S. Travis
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-2619
• Calibration of User Equilibrium Model with Heterogeneous Risk Attitudes
  Abstract: We calibrate a user equilibrium model with heterogeneous risk attitudes using laboratory data collected from human subjects making repeated route choices and collectively generating congestion in a hypothetical network subject to random disruptions. A traveler is assumed to take the minimum expected disutility route, where the disutility is a power function of the travel time and the value of the power determines the risk attitude. A multi-class user equilibrium model is developed where travelers have different risk attitudes across classes. The discretized probability distribution of the risk attitude is calibrated by minimizing the deviation of the predicted route flows from the observed route flows averaged over a relative steady period. A crude disaggregate analysis is conducted for each individual so that constraints on the risk attitude distribution are derived and the search space is reduced. Uniform distributions with parameters that differ in the risk averse and seeking ranges are used. Most subjects seem to be risk seeking, which might be due to the recency effect found in psychology literature that human beings put more weights on recent outcomes and thus frequencies of rare events (incidents that reduce road capacities in the experiment) are underestimated.
  Authors: Pothering, Ryan Joseph; Gao, Song
  Authors: Pothering, Ryan Joseph; Gao, Song
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-3641
• Effect of Delay Frequency and Duration as a Measure of Reliability on Route Choice
  Abstract: The importance of economic value of travel time reliability and behavioral responses to it are now well recognized in the transportation literature. To justify and incorporate this measure into decision making, public agencies are now more inclined to assess the use of value of travel time reliability in project evaluations. We aim to aid that process by measuring travelers’ behavioral responses to travel time reliability and their willingness to pay (WTP) to avoid unreliable routes. Travelers’ route choice attitudes are measured through a pivoted stated preference (SP) survey technique. The contribution of this study lies in our effort to express travel time reliability based on two dimensions: the frequency of unexpected delay, and the magnitude of unexpected delay. We simply compare days with unexpected delay to days with usual travel time, thereby circumventing the issue of presenting numerical distributions and statistical terms to day-to-day commuters. A mixed logit framework was used to understand travelers’ route choice attitudes. We found that it’s not only the frequency but also the amount of unexpected delay attached to it that predicts the preference over a particular route. The study also contributes to the existing literature by calculating WTP measures that correspond to different levels of frequency of unexpected delay days. We calculated three different WTP measures, including the WTP for travel time, the WTP for frequency embedded travel time, and the WTP for travel time reliability.
  Authors: Sikka, Nikhil; Hanley, Paul F.
  Authors: Sikka, Nikhil; Hanley, Paul F.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-3091
• Process Model for Route Choice in Risky Traffic Networks
  Abstract: Unreliable travel times are a major obstacle to an efficient and reliable transportation system. The accurate modeling of travelers' route choice decision making when faced with unreliable (risky) travel times is necessary for the assessment of policies aimed at improving travel time reliability. Compared with econometric models, process models have not been investigated in travel decision making under risk. A process model aims to describe the actual decision making procedure and could potentially provide a better explanation to route choice behavior. A process model, Priority Heuristic (PH) model,was introduced and its probabilistic version, Probabilistic Priority Heuristic (PPH) model, was developed in this study. With data collected from a stated preference (SP) survey, one econometric model, Rank-Dependent Expected Utility (RDEU) model, and two other alternative models were compared in a cross validation to investigate their data-fitting and predictive performance. Final results suggested that PPH model outperformed RDEU in both data-fitting and predictive performance and could be a promising new area in travel behavior research.
  Authors: Tian, Hengliang; Gao, Song
  Authors: Tian, Hengliang; Gao, Song
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-4656
• Optimal Adaptive Routing Algorithm for Large-Scale Stochastic Time-Dependent Networks
  Abstract: The objective of the research is to study optimal routing policy (ORP) problems and to develop an ORP algorithm practical for large real-life networks, where a traveler could revise the route choice based upon en route information. The routing problems studied can be viewed as counterparts of shortest path problems in deterministic networks. A routing policy is defined as a decision rule that specifies what node to take next at each decision node based on realized link travel times and the current time. The existing routing policy algorithm is for explorative purpose and can only be applied to hypothetical simplified network. In this research, important changes have been made to make it practical in a real-life network. Important changes in the new algorithm include piece-wise linear travel time representation, turn-based, label-correcting, and dynamic blocked links. We then study complete dependency perfect online information (CDPI) variant in a real-life network (Pioneer Valley, Massachusetts). Link travel times are modeled as random variables with time-dependent distributions which are obtained by running Dynamic Traffic Assignment (DTA) using data provided by Pioneer Valley Planning Commission (PVPC). We give a comprehensive explanation of the changes by comparing the two algorithms and an in-depth discussion of the parameters that affects the running time of the new algorithm. We then carry out computational tests on the running times changing with different parameters and present the summary of its effectiveness. This research is also a good starting point to investigate strategic route choice models and strategic route choice behavior in a real-life network. The major tasks are to acquire data, generate time-adaptive routing policies, and estimate running time of the algorithm by changing the parameters. The research contributes to the knowledge base of ORP problems in stochastic time-dependent (STD) networks by developing an algorithm practical for large-scale networks that considers complete time-wise and link-wise stochastic dependency.
  Authors: Ding, Jing; Gao, Song
  Authors: Ding, Jing; Gao, Song
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-4273
• Snowball Effect and Traffic Equilibrium in a Market Entry Game: Laboratory Experiment
  Abstract: The Market Entry Game (Selten \& Guth, 1982 ; Gary-Bobo, 1990) is a coordination game where average cost of entry increase linearly. Following the suggestion of Anderson et al. (2008) and empirical evidence in transport economics suggested by related literature about the Speed-Flow Relationship (Verhoef, 2005), we built a theoretical model of market entry game where average entry cost grows at an increasing rate when the flow of entrants rises (snowball MEG). The calibration of experimental parameters enables us to compare outcomes regarding Snowball MEG to a benchmark based on a usual linear MEG, its experimental calibration being based on Anderson et al., 2008). Then, we run an experiment where theoretical predictions for both models give comparable aggregate entry rate and exhibit common properties. Moreover, we conduct variation about group size in order to assess size effect on coordination level. Basically, we find no significant difference between the aggregated entry rates in our snowball game compared to usual MEG. But size clearly matters, especially for issue regarding successful coordination for users in the snowball MEG.
  Authors: Denant-Boemont, Laurent; Fortat, Vivien
  Authors: Denant-Boemont, Laurent; Fortat, Vivien
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-1103
  Practice-Ready: Yes
• Braess Paradox Under Boundedly Rational User Equilibria: Properties on the Paradox Network
  Abstract: The classical Braess' Paradox was built upon a perfectly rational behavioral assumption. In the literature, many empirical studies have demonstrated that perfect rationality is too restrictive to happen in reality. Recently, a number of transportation network design models have been established by replacing the perfect rationality with bounded rationality, in which an indifference bound is introduced to indicate the upper bound of non-shortest paths that drivers are willing to take. However, the analytical properties of Braess' Paradox under bounded rationality remain unanswered. This paper aims at filling in the gap by exploring the relationships between the occurrence of Braess' Paradox and the indifference bound in bounded rationality as well as demand level. We use the classical Braess' Paradox network to derive these relationships based upon the worst flow pattern in the bounded rational user equilibria set, since transportation planners generally tend to be risk-averse and attempt to improve the network performance under the worst traffic condition. The unveiled relationships offer a guideline for urban planners to prevent the occurrence of Braess' Paradox.
  Authors: Di, Xuan; He, Xiaozheng; Liu, Henry X.
  Authors: Di, Xuan; He, Xiaozheng; Liu, Henry X.
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-3296
  Practice-Ready: Yes
• Using a DTA Model to Evaluate Road Tolling Strategies: Seattle Experience
  Abstract: Recent applications and research in toll modeling suggest that Dynamic Traffic Assignment (DTA) modeling is gaining acceptance in the industry as a suitable modeling platform for a variety of planning studies and projects. By blending the traffic assignment capabilities of travel demand modeling tools with the intersection/link analysis capacity of traffic simulation tools, DTA modeling effectively bridges the “gap” between the more commonly used macroscopic and microscopic paradigms. Despite increasing acceptance and use, however, DTA modeling for the purposes of tolling analysis and development of toll revenue projections has been limited, even though the benefits and effectiveness of such applications appear promising. This paper summarizes a current effort to develop a DTA model—which incorporates multiple user classes—to support updates to the tolling and revenue analysis for the Alaskan Way Viaduct Replacement Tunnel Project in Seattle, Washington. The primary objectives and motivations for using DTA are first described and are followed by an overview of the study methodology and tolling model development process. Results and findings of the study are summarized, including lessons learned. Future potential applications of the model are then discussed, along with how this work will potentially add value to toll modeling/forecasting.
  Authors: Dehghani, Youssef
  Authors: Dehghani, Youssef
  Year: 2013
  Document Type: Paper
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-4652
  Practice-Ready: Yes
• Snowball Effect and Traffic Equilibrium in a Market Entry Game: Laboratory Experiment
  Authors: Denant-Boemont, Laurent
  Authors: Denant-Boemont, Laurent
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-1103
• An Alpha-Reliable Combined Mean Travel Time Measure for Modeling Routing Risk Behaviors in Stochastic Traffic Network
  Authors: Zhang, Wenyi
  Authors: Zhang, Wenyi
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-0611
• Latent Class Choice Model of Heterogeneous Drivers' Route Choice Behavior Based on Real-World Experiment
  Authors: Rakha, Hesham
  Authors: Rakha, Hesham
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-1367
• Latent Class Choice Model of Heterogeneous Drivers' Route Choice Behavior Based on Real-World Experiment
  Authors: Tawfik, Aly
  Authors: Tawfik, Aly
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-1367
• Modeling Combined Travel Choices of Electric Vehicle Drivers with a Variational Inequality Network Formulation
  Authors: Zhang, Ti
  Authors: Zhang, Ti
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-2619
• Effect of Delay Frequency and Duration as a Measure of Reliability on Route Choice
  Authors: Sikka, Nikhil
  Authors: Sikka, Nikhil
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-3091
• Modeling Framework to Analyze Effect of Multiple Traffic Information Service Providers on Traffic Network Performance
  Authors: Yang, Inchul
  Authors: Yang, Inchul
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-2392
• Calibration of User Equilibrium Model with Heterogeneous Risk Attitudes
  Authors: Pothering, Ryan
  Authors: Pothering, Ryan
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-3641
• Process Model for Route Choice in Risky Traffic Networks
  Authors: Tian, Hengliang
  Authors: Tian, Hengliang
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-4656
• Braess Paradox Under Boundedly Rational User Equilibria: Properties on the Paradox Network
  Authors: Di, Xuan
  Authors: Di, Xuan
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-3296
• Optimal Adaptive Routing Algorithm for Large-Scale Stochastic Time-Dependent Networks
  Authors: Ding, Jing
  Authors: Ding, Jing
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-4273
• Using a DTA Model to Evaluate Road Tolling Strategies: Seattle Experience
  Authors: Dehghani, Youssef
  Authors: Dehghani, Youssef
  Year: 2013
  Document Type: Presentation
  Subject: Planning and Forecasting
  Session: 688
  Paper Number: 13-4652