Density-based clustering methodology for estimating fuel consumption of intracity bus by using DTG data

Abstract

This study suggests the methodology of estimating the standard bus fuel consumption based on large scale data. The purpose of the estimation is to encourage private transport companies to decrease the fuel costs of their intracity buses. To calculate the standard fuel consumption, we used the digital tachograph data of all intracity buses, and applied the density-based clustering analysis method to systematically eliminate many outliers which might be caused by system errors. With the suggested methodology, we calculated the accurate driving distance of each bus route, analyzed it with the bus fueling data, and estimated the monthly standard fuel consumption of each bus route. Through this study, it is expected to significantly improve the accuracy of the standard fuel consumption of intracity buses by removing the errors in the data.