Tomer Toledo

Modeling Integrated Lane-changing Behavior

The lane-changing model is an important component within microscopic traffic simulation tools. Following the emergence of these tools in recent years, interest in the development of more reliable lane-changing models has increased. Lane-changing behavior is also important in several other applications such as capacity analysis and safety studies. Lane-changing behavior is usually modeled in two steps: (a) the decision to consider a lane change, and (b) the decision to execute the lane change. In most models, lane changes are classified as either mandatory (MLC) or discretionary (DLC). MLC are performed when the driver must leave the current lane. DLC are performed to improve driving conditions. Gap acceptance models are used to model the execution of lane changes. The classification of lane changes as either mandatory or discretionary prohibits capturing trade-offs between these considerations. The result is a rigid behavioral structure that does not permit, for example, overtaking when mandatory considerations are active. Using these models within a microsimulator may result in unrealistic traffic flow characteristics. In addition, little empirical work has been done to rigorously estimate the parameters of lane-changing models. An integrated lane-changing model, which allows drivers to jointly consider mandatory and discretionary considerations, is presented. Parameters of the model are estimated with detailed vehicle trajectory data.

Modeling the behavior of novice young drivers during the first year after licensure

Novice young drivers suffer from increased crash risk that translates into over-representation in road injuries. In order to effectively confront this problem, a better understanding of the driving behavior of novice young drivers and of its determinants is needed. This study analyzes the behavior of novice young drivers within a Graduated Driver Licensing (GDL) program. Data on driving behavior of 62 novice drivers and their parents, who voluntarily participated in this experiment, were collected using in-vehicle data recorders that calculate compound risk indices as measures of the risk taking behavior of drivers. Data were used to estimate a negative binomial model to identify major determinants that affect the driving behavior of young drivers during the first year after licensure. Estimation results suggest that the risk taking behavior of young drivers is influenced by gender, sensation seeking tendency, driving behavior of their parents, amount of supervised driving and level of parental monitoring.

CALIBRATION AND VALIDATION OF MICROSCOPIC TRAFFIC SIMULATION TOOLS: STOCKHOLM CASE STUDY

The calibration and validation approach and results from a case study applying the microscopic traffic simulation tool MITSIMLab to a mixed urban-freeway network in the Brunnsviken area in the north of Stockholm, Sweden, under congested traffic conditions are described. Two important components of the simulator were calibrated: driving behavior models and travel behavior components, including origin–destination flows and the route choice model. In the absence of detailed data, only aggregate data (i.e., speed and flow measurements at sensor locations) were available for calibration. Aggregate calibration uses simulation output, which is a result of the interaction among all components of the simulator. Therefore, it is, in general, impossible to identify the effect of individual models on traffic flow when using aggregate data. The calibration approach used takes these interactions into account by iteratively calibrating the different components to minimize the deviation between observed and simulated measurements. The calibrated MITSIMLab model was validated by comparing observed and simulated measurements: traffic flows at sensor locations, point-to-point travel times, and queue lengths. A second set of measurements, taken a year after the ones used for calibration, was used at this stage. Results of the validation are presented. Practical difficulties and limitations that may arise with application of the calibration and validation approach are discussed.

Driving Behaviour: Models and Challenges

Abstract Driving behaviour models capture drivers’ tactical manoeuvring decisions in different traffic conditions. These models are essential to microscopic traffic simulation systems. The paper reviews the state‐of‐the‐art in the main areas of driving behaviour research: acceleration, lane changing and gap acceptance. Overall, the main limitation of current models is that in many cases they do not adequately capture the sophistication of drivers: they do not capture the interdependencies among the decisions made by the same drivers over time and across decision dimensions; they represent instantaneous decision‐making, which fails to capture drivers’ planning and anticipation capabilities; and only capture myopic considerations that do not account for extended driving goals and considerations. Furthermore, most models proposed in the literature were not estimated rigorously. In many cases, this is due to the limited availability of detailed trajectory data, which are required for estimation. Hence, data availability poses a significant obstacle to the advancement of driving behaviour modelling.

CALIBRATION OF MICROSCOPIC TRAFFIC SIMULATION MODELS WITH AGGREGATE DATA

A framework for the calibration of microscopic traffic simulation models using aggregate data is presented. The framework takes into account the interactions between the various inputs and parameters of the simulator by estimating origin-destination (O-D) flows jointly with the behavioral parameters. An optimization-based approach is used for the joint calibration. Since the calibration of the parameters depends on the estimated O-D flows and vice versa, the proposed framework is iterative. O-D estimation is based on the well-known generalized least squares estimator. A systematic search approach based on the complex algorithm is adopted for calibration of the behavioral parameters. This algorithm is particularly useful for the problem at hand since it does not require calculations of derivatives of the objective function. The applicability of the approach is demonstrated through its application to case studies using MITSIMLab, a microscopic traffic simulation model.

In-Vehicle Data Recorder for Evaluation of Driving Behavior and Safety

This paper describes the overall framework and components of an in-vehicle data recorder (IVDR) called DriveDiagnostics and presents results from a study to validate its performance. This IVDR has been designed to monitor and analyze driver behavior not only in crash or precrash events but also in normal driving situations. It records the movement of the vehicle and uses this information to indicate overall trip safety. A validation study involved 33 drivers whose vehicles were instrumented with the IVDR. The experiment first included a blind profiling stage in which drivers did not receive any feedback from the system; that stage was followed by a feedback stage in which drivers had access to personal web pages with the information recorded by the system. Data collected in the blind profiling stage was used to investigate the connection between driver safety indices as captured by the system and historic crash data. The results show significant correlations between the two data sets, suggesting that the ...

EFFECTS OF CHOICE SET SIZE AND ROUTE CHOICE MODELS ON PATH-BASED TRAFFIC ASSIGNMENT

Few of the recently developed route choice models have actually been applied in traffic assignment problems. This paper discusses the implementation of selected route choice models in stochastic user equilibrium algorithms. The focus of the paper is on path-based assignment, which is essential in the implementation of route choice models. The paper analyzes the effect of choice set size and selected choice models on problem convergence, running time and selected results. The results presented in the paper indicate that for real-size networks, generation of a large number of alternative routes is needed. Furthermore, convergence properties greatly improve if the generated routes are sufficiently disjointed.

Evaluation of the Potential Benefits of Advanced Traveler Information Systems

This article evaluates the potential travel time savings from Advanced Traveler Information Systems (ATIS) that provide drivers with travel time and routing information. We classify ATIS in various levels based on the type of information they use to generate guidance and the timing of the dissemination of the generated guidance to drivers. We present a case study that examines the potential travel time savings of ATIS as well as the implications on travel time variability and reliability and the sensitivity of the results to the accuracy of the information, using real-world data collected from a freeway network in Los Angeles, California.

Estimation of Vehicle Trajectories with Locally Weighted Regression

Vehicle trajectory data are important for calibrating driver behavior models (e.g., car following, acceleration, lane changing, and gap acceptance). The data are usually collected through imaging technologies, such as video. Processing these data may require substantial effort, and the resulting trajectories usually contain measurement and processing errors while also missing data points. An approach is presented to the processing of position data to develop vehicle trajectories and consequently speed and acceleration profiles. The approach uses local regression, a method well suited for mapping highly nonlinear functions. The proposed methodology is applied to a set of position data. The results demonstrate the value of the method to development of vehicle trajectories and speed and acceleration profiles. The conducted sensitivity analysis also shows that the method is rather robust regarding measurement errors and missing values.

Statistical Validation of Traffic Simulation Models

Traffic simulation models support detailed analysis of the dynamics of traffic phenomena and are important tools for analysis of transportation systems. In order to evaluate correctly the impact of different traffic management schemes, simulation models must be able to replicate reality adequately. Model validation (i.e., the process of checking to what extent the model replicates reality) is discussed. The role of validation is defined within the scope of model development and calibration, and the framework for performing the validation is discussed. A hierarchy of statistical methods to validate different types of simulation outputs against observed data is examined. Also, a validation method is proposed on the basis of statistical tests on metamodels fitted to the observed and simulated data. A case study illustrates the applicability of the various methods.