Work Zone Characteristics in Microscopic Acceleration Modeling: Calibration and Numerical Analysis Open Access
Downloadable ContentDownload PDF
The essential need to maintain roadway structures, makes encountering work zones an inevitable part of today's driving. Presence of work zones can impact the traffic flow characteristics in the macroscopic and the microscopic levels. The corresponding changes in traffic dynamics directly affect the mobility and the safety in work zone areas. As the work zone changes in terms of layout and design, the driver behavior changes. Accordingly, a closer look into driver's interaction with work zones geometrics and activity features may be helpful. The objective of this study is to examine the longitudinal characteristics of driver behavior in different work zone conditions using a driving simulator. For this purpose, a detailed literature review motivated the examination of work zone related factors. The corresponding 3 exogenous variables are the work zone length, the barrier type (between the activity area and the passing traffic) and the level of activity. 8 different scenarios were designed by changing these 3 variables in the STISIM driving simulator. After implementing the driving tests, the output trajectory data is recorded for each subject driver (i.e., participant).In order to quantify the driver's perception of the surrounding environment (i.e., work zone layout and activity) and his/her judgment-execution process, a prospect-theory (PT) based acceleration model is adopted in this study. This model translates a subjective utility-maximization framework to longitudinal acceleration maneuvering. In other words, the driver's longitudinal behavior is modeled using a microscopic traffic acceleration model. The parameters of the corresponding PT model are calibrated using genetic algorithm. Exploring the differences in parameter values for different scenarios helps understanding the impacts that one or a combination of external factors may have on longitudinal driver behavior in work zone areas. Initial analysis of the driver's acceleration behavior shows that drivers tend to attain higher average speeds and shorter average space headways in work zone areas characterized by jersey barriers. It also showed that drivers attain higher speeds at long stretches of work zones and in low-level activity work zones. After calibrating the parameters of the above mentioned PT model for each scenario, a series of multivariate analysis of variance (MANOVA) tests were conducted to examine the validity of the assumption that differences observed in the calibrated parameters are caused by the external variables (i.e., work zone characteristics) rather than inter-driver heterogeneity. The test results showed that different parameters are sensitive to different work zone characteristics. The parameter (indicator of the intensity of the driver's reaction to external factors) changes most when the level of activity changes which means drivers react to changes in level of activity more extremely. Driver's evaluation of the relative losses in velocity with respect to the relative gains (i.e., parameter ) is more sensitive to the changes in barrier type and activity level indicating more or less risk averseness. Finally, the drivers uncertainty about the leader's relative speed (i.e., parameter ) and driver's reaction time (parameter RT) are affected the most when changes in the work zone activity level are observed.