Alexandra Kondyli

Modeling Driver Behavior at Freeway-Ramp Merges

Freeway–ramp merging segments are important components of freeway facilities. The composite behavior of acceleration and gap acceptance of the merging traffic as well as the cooperative behavior of the freeway traffic can result in conflicts and trigger congestion. The goal of this research was to develop a ramp-merging model that considered the merging process as perceived by drivers and to investigate the contribution of individual drivers’ merging behavior to the breakdown event. Focus group meetings were conducted, and drivers’ merging behavior was observed in in-vehicle experiments. The data were used to develop a gap-acceptance model under different merging conditions and a model of driver behavior that predicted vehicle interactions on the freeway with merging vehicles, considering different driver types. A merging turbulence model that evaluated the effect of vehicle interactions on traffic flow also is presented. Study findings can be used to refine existing micro-simulation models, realistically replicate freeway flow breakdown, and provide insight into the triggers of the breakdown of freeway flow.

Estimation of Freeway Work Zone Capacity Through Simulation and Field Data

The objective of this research was to develop analytical models and procedures for estimating the capacity of a freeway work zone by considering various geometric-, traffic-, and work zone–related parameters. The study was conducted in two stages: simulation-based modeling and field data collection. In the first stage, CORSIM (Version 5.1) was used to develop a comprehensive database for various work zone scenarios. Analytical models were developed to predict work zone capacity on the basis of these simulated data and previous literature findings considering three work zone configurations: two-to-one, three-to-two, and three-to-one lane closures. In the second stage, field data were collected at a freeway work zone to evaluate and refine the analytical models developed. Data were collected at the freeway work zone site during 15 evening peak periods, which included left- and right-lane closures as well as rainy weather conditions. The observed capacities were compared with those predicted by the new analytical models as well as to those estimated by the Highway Capacity Manual 2000. It was concluded that the analytical models developed predicted within 1% the capacity of the study work zone.

Assessment of car-following models by driver type and under different traffic, weather conditions using data from an instrumented vehicle

Abstract Car-following models are important components of simulation tools, since they describe the behavior of the following vehicle as a function of the lead vehicle trajectory. Several models have been developed and evaluated using field data. However, the literature has been inconclusive regarding the applicability of various car-following models under different operational conditions such as congested vs. non-congested. There has been very limited research regarding the relationship between car-following calibration parameters and different driver types. The objective of this study was to assess four car-following models using field data under different traffic (congested vs. uncongested) and weather conditions (rain vs. clear sky) and for various driver types (aggressive, average, and conservative). The assessed models were the Gipps (component of the AIMSUN software), Pitt (component of the CORSIM software), MITSIM (utilized in MITSIMLab program), and the Modified Pitt model. The data used in the analysis were collected with the help of an instrumented vehicle. The field trajectories were compared to the trajectories obtained by each of the four models evaluated. Results showed that the variable predicted best by the models was the speed of the following vehicle, which is consistent with previous findings. The calibration analysis also showed that the best variable to be used for calibration is spacing. Calibrating by spacing minimizes the errors that can be accumulated and can distort the final trajectory. Three calibration analyses were completed: first using all data available, second by traffic condition, and third by driver type. The best results were obtained when the parameters were calibrated by driver type using the MITSIM model. The study concludes with recommended calibration parameters, and application guidelines related to the car-following models examined.

Driver Behavior at Freeway-Ramp Merging Areas: Focus Group Findings

Driver behavior is an important factor affecting traffic operations, and it accounts to a large extent for the observed variability in roadway performance. Researchers have developed various driver behavior–related models for operational analysis and simulation applications. These models consider drivers’ choices related to gap identification and selection, lane changing, and so on. However, the literature related to drivers’ thinking processes and support for the assumptions used in existing models is scarce. This study conducted three focus groups to investigate drivers’ intended actions along a freeway-ramp merging segment. Several scenarios were discussed in which participants indicated their thinking process and likely actions while merging or traversing a merging segment. The study considers noncongested and congested traffic conditions. It also correlates the drivers’ responses to their individual characteristics.

Development and Evaluation of Methods for Constructing Breakdown Probability Models

AbstractThe term breakdown of flow at a freeway bottleneck location has been used to describe the transition from relatively free-flowing traffic to congestion, often called stop-and-go traffic, but more generally experienced as slow-and-go. The focus of the research presented in this paper was to develop probabilistic models to predict breakdown of flow by using data from five freeway-ramp merging segments. The development of the probabilistic models allows exploration of several issues: the identification of breakdown, the performance measures for the identification of the breakdown, and the breakdown location. The breakdown probability model (BPM) is based on lifetime data analysis statistics. The breakdown probability models can be very useful tools in establishing capacity values, which in turn can be part of the freeway traffic management process.

Sensitivity analysis of CORSIM with respect to the process of freeway flow breakdown at bottleneck locations

Abstract Various microscopic simulation models have been used for studying traffic operations along freeway segments. An important desirable function of these models is their ability to obtain capacity and replicate the breakdown process realistically. The objectives of this paper are to evaluate the capability of a microsimulation model, CORSIM, to replicate the process of breakdown and to perform a sensitivity analysis on driver behavior parameters. The research findings indicate that CORSIM has some strengths and some weaknesses with respect to the breakdown process. Sensitivity analysis shows the different effects of these parameters on the breakdown occurrence and provides recommendations on the application of these parameters to provide a more realistic representation of traffic operations.

Exploratory Analysis of Lane Changing on Freeways Based on Driver Behavior

AbstractLane changing has received much attention as it is a significant component of microscopic traffic simulation. Many studies have focused on the details of the lane changing maneuver from external observation-based data which do not consider the type of driver performing the maneuver. The research reported in this paper relates the physical details of freeway lane changing to the type of driver performing the maneuver. Forty-six research participants drove an instrumented vehicle and performed a combined total of 726 freeway lane changes. Each research participant was categorized into one of four groups ranging from conservative to aggressive based on cluster analysis. The data were analyzed to identify any trends between the different driver types and their lane changing characteristics, specifically lane changing duration and gap acceptance characteristics. In general, more conservative drivers have greater lane changing durations than aggressive drivers. The gap acceptance comparison among driver...

Freeway Capacity Estimation Method for Planning Applications

The capacity of a freeway segment is a critical factor for the planning, design, and analysis of freeway facilities. The Highway Capacity Manual (HCM) is considered to be one of the authoritative sources on capacity values for a variety of roadway types in the United States, particularly for planning purposes. For basic freeway segments, a single set of capacity values are provided as a function of free-flow speed. Although these values are considered to be reasonably representative values for freeways located throughout the United States, the HCM does not provide any guidance on how its recommended values can be adjusted to reflect significant differences in capacity due to local conditions, nor how to directly measure or estimate capacity values. With the recent development of a statewide freeway traffic data archive in Florida, the Florida Department of Transportation (FDOT) desired to use these data to determine if the freeway capacity values in the HCM were appropriate for Florida freeway level of service analyses for planning and preliminary engineering applications. This study evaluated two previously published methods for estimating freeway capacity, identified their advantages and disadvantages for use in planning applications, and ultimately recommended a method for use by FDOT. One of the chosen methods determines capacity from a mathematical function generated from speed-flow data points and the other chosen method uses the concept of breakdown probability distribution to determine capacity. Neither of these methods proved desirable for planning applications due to computational burden; thus, an alternate, simplified approach was developed that used a simple averaging method of highest flow rates to determine capacity.

An Empirical Framework for Intersection Optimization Based on Uniform Design

Operational performance optimization of signalized intersections is one of the most important tasks for traffic engineers and researchers. To compensate for the limitations of practical implementation, simulation software packages have been widely used to evaluate different optimization strategies and thus to improve the efficiency of the intersections as well as the entire network. However, for the existing optimization studies on signalized intersections, the relationships among various optimization measures and the combination of strategies have not been fully investigated. In this paper, uniform design experimentation was introduced to combine different optimization measures into strategies and achieve the minimum time cost in model construction. VISSIM software package was then calibrated and used to evaluate various optimization strategies and identify the one with the best measurement of performance, namely, control delay at the signalized intersection. By taking a representative congested intersection in Shanghai as a case study, the optimal strategy was identified to reduce the overall control delay by 27.3%, which further verified the modeling capability of the proposed method.

Comparison of Highway Capacity Estimation Methods

The capacity of a freeway segment is a critical factor for planning, design, and operational analysis of freeway facilities. This research aimed to perform a comparison among well-known freeway capacity estimation methods in order to investigate their application, as well as their advantages and disadvantages. Single estimate capacity methods such as the Van Aerde method, and breakdown probability methods, such as the product limit method (PLM), the Highway Capacity Manual (HCM) method, and the sustainable flow index (SFI) method, were applied at six merge bottleneck locations in the Kansas City area. The results from all methods were compared and the advantages and the disadvantages of each method were discussed. The HCM results showed a significant variability in the estimated breakdown probability function and the resulting capacities. The HCM method was also found to be sensitive to the breakdown probability ratios as a single breakdown observation can significantly shift the fitted distribution and the corresponding capacity estimate. The PLM model provided the highest capacity estimates, followed by the Van Aerde model capacities. The Van Aerde capacities were also found to be closer to the average pre-breakdown flow rates. Finally, the PLM and the SFI method showed consistent performance in comparison to the remaining methods, and flexibility in being applied on different sites with various characteristics.