Speed distributions at a controlled junction using microscopic real data from image processing
van Arem, B.
TNO Bouw en Ondergrond
Modeling realistic driving behavior at signalized junctions is fundamental for many applications, for instance for determining the signal capacity or for estimating traffic emissions. For these measures, the knowledge of speed, acceleration and deceleration of individual vehicles passing the signal is crucial. This paper analyses these characteristics by using real data collected using image processing and clustering techniques. These methodologies allow one to obtain vehicle trajectories with very high detail, providing individual vehicles speeds at a microscopic level. In this paper we focus on analyzing the empirical distributions of speeds observed with this unique dataset near and up to a few meters upstream the stop-sign. We compared these distributions with the results of repeated simulations of two microscopic software programs. Some inconsistencies were found with this comparison. A direct consequence of these inconsistencies might be in the estimation of traffic emissions, which are expected to be underestimated if microscopic simulation programs are not well calibrated.
To reference this document use:
Proceedings of the 11th International IEEE Conference on Intelligent Transportation Systems (ITSC2008), October 12-15, 2008, Beijing, China, 6