Utilizing Public Vehicle Travel Survey Data Sets for Vehicle Driving Pattern and Fuel Economy Studies

摘要

Realistic vehicle fuel economy studies require real-world vehicle driving behavior data along with various factors affecting the fuel consumption. Such studies require data with various vehicles usages for prolonged periods of time. A project dedicated to collecting such data is an enormous and costly undertaking. Alternatively, we propose to utilize two publicly available vehicle travel survey data sets. One is Puget Sound Travel Survey collected using GPS devices in 484 vehicles between 2004 and 2006. Over 750,000 trips were recorded with a 10-second time resolution. The data were obtained to study travel behavior changes in response to time-and-location-variable road tolling. The other is Atlanta Regional Commission Travel Survey conducted for a comprehensive study of the demographic and travel behavior characteristics of residents within the study area. The data was collected every second using GPS devices in vehicles during a two-month period (March-May and July-September 2011) with a maximum of seven days recording. A total of 1,653 vehicles participated with about 40,000 trips. The relatively high time resolution makes the data set useful for second-by-second analysis. These data sets are complementary in terms of time resolution and a range of traffic and environmental conditions. We have analyzed the data. Our findings, including a variety of characteristics of the data sets (e.g., speed and distance distributions of the trips), are presented. We conclude that the two complementary data sources can be combined to form a new and representative data set suitable for studying real-world vehicle driving pattern and fuel economy research.