Date of Award
Doctor of Philosophy
Civil, Construction, and Environmental Engineering
Shauna L. Hallmark
Transportation is a major source of many major air pollutants as well as greenhouse gas emissions. The four common factors responsible for vehicular emissions are vehicle, road characteristics, traffic conditions and driving behavior. The objective of this dissertation was to study driving behavior since it is highly correlated to emissions as shown by previous studies. Understanding driving behavior is likely to help improve emissions estimates. In this dissertation, three levels of analyses of driving behavior were conducted including: (1) exploring driving behavior parameters and assessing their impact on emissions, (2) comparing driving behavior among the three most common traffic control devices, and (3) modeling second-by-second driving behavior of individual drivers. In order to explore these relationships, spatial location, vehicle kinematics, and CO2 emissions were collected along a study road corridor in Urbandale (IA) was. The chosen road corridor comprised of a roundabout, an all-way-stop and a traffic signal along with curve and tangent sections. The traffic during peak and off-peak hours on the corridor was comparable. This was useful for comparing driving behavior across drivers under similar conditions. A single instrumented vehicle was driven over the corridor by four different subject drivers. The vehicle was equipped with a portable emissions measurement device which had engine sensor, tail-pipe sample lines and a GPS.
In the first analysis, vehicle kinematic variables were used to derive driving behavior parameters that included gas pedal use and brake pedal use. Two groups of drivers were identified based on these parameters. The study identified gaspad and brakepad as important driving behavior parameters which can explain variation in vehicular emissions.
Driving behavior parameters used in previous studies for developing driving cycle were utilized in this study to compare driving behavior between traffic control devices for the second analysis. These parameters characterized speed behavior, speed change behavior and energy gain behavior. A MANOVA model was used for comparing the overall driving behavior between traffic control devices by comparing these parameters. Results showed that driving behavior at the roundabout and all-way-stop differ significantly (p < 0.001) on at least one of driving behavior parameter. Likewise, roundabout and traffic signals also differed in terms of driving behavior (p < 0.001). Driving behavior and emissions are highly correlated. This implies using separate emission factors for different traffic control devices.
In the third analysis, speed profiles at roundabout were modeled for the drivers using a fourth degree polynomial regression. Results showed that speed profiles models were significantly different across drivers. This implied that drivers must be treated as random variables in modeling driving behavior and emissions for a given road or driver population. Average speeds of drivers at yield point were simulated based on the model. The maximum difference was found to be about 1.5 mph.
Mudgal, Abhisek, "Modeling Driving Behavior at Traffic Control Devices" (2011). Graduate Theses and Dissertations. 10343.