摘要

The main contribution of this paper is to show, on the basis of empirical traffic data, that the distribution of vehicle speeds on a road segment evolves with the occupancy in a simple manner. Under a critical occupancy, the distribution is unimodal, with one peak (the high mode) close to the free-flow speed. When the occupancy exceeds the critical occupancy, the distribution of speeds becomes bimodal. A second peak (the low mode) appears, at a noticeably lower speed than the high mode. Empirical speed distributions are well fitted when assuming a (appropriately scaled and translated) low temperature Maxwell-Boltzmann distribution for the high mode and a high-temperature distribution for the low mode. The standard fundamental diagram expresses the mean flow speed as a function of occupancy. The model proposed in this paper expresses the distribution of vehicle speeds as a function of the occupancy. We believe this result to be of great importance for both the theory of traffic flow and practical applications.

  • 出版日期2014-3