| Speed and acceleration vehicle distributions under real-world driving conditions are required as input to improved motor vehicle emissions models, and are being used in other environmental and fuel economy models. Accurate and representative measurements of these real-world conditions are necessary for development of predictive and simulation models, and to better understand and model driver behavior. The speeds and accelerations of vehicles will influence the power demand on the vehicle (along with other variables such as air conditions, towing, vehicle weight, etc.) which in turn influences emissions rates.;The goal of this research is to develop a procedure for collection of vehicle modal activity, and to develop models to predict the percentage of activity in a given speed-acceleration range based on roadway, traffic stream, and vehicle fleet characteristics. After developing new software, laser rangefinders were used to remotely measure vehicle modal activity. The new system allows users to determine modal operating differences across different vehicle classifications. Vehicle activity data used to develop the models represents 27 freeway segment locations in the Atlanta, Georgia region (6,257 vehicles and 48,785 seconds of activity).;The distributions of modal activity are statistically related to the physical and operational characteristics of the transportation network. Ordinary least squares regression models predict the percent of activity based on inputs of roadway geometry, vehicle type, and traffic conditions. Results show that roadway geometry and traffic conditions significantly affect operating mode fractions, factors that are currently not taken into account in transportation and emissions models. Geometric variables in different models include lane number, lane width, grade, and/or curvature. Operational variables significant in the model include density, flow of traffic and/or average speed. Models for the different modal properties (acceleration, deceleration, power, and PKE) include different combinations of the variables to predict the percent of activity. |
