A Parsimonious Method for Offline Freeway Travel Time Estimation From Sectional Speed Detectors

Abstract

Online travel time estimation is receiving increasing attention. However, offline algorithms remain important, as databases of historical travel times are needed for many intelligent transportation systems applications. In this paper a parsimonious method is presented for estimating off-line travel times based on local speed measurements along a route with uninterrupted flow, for example, on motorways. The interpolation between the different detector locations is done by reconstructing trajectories based on traffic flow theoretical principles. The proposed method is efficient, easy to calibrate, and shows reliable results even when detector spacing is relatively high. The method is compared to other existing methods and validated on a 40-km highway.