Files

Abstract

Modeling traffic flow has historically been tackled using a variety of different approaches. The microscopic method uses a system of ordinary differential equations to model the interaction of a single car with its surroundings. This model has fixed pa- rameter values for all driver behavior characteristics, such as reaction time and reaction strength. Given that these parameters represent driver behavior, it is more realistic to introduce some stochasticity into the model parameters and observe the effect on the resulting traffic conditions. In this paper, direct simulation results show that stochastically varied parameter values produce larger free-flow regions, faster stabilization to free-flow conditions, and reduced traffic intensity overall. Fundamental diagrams show that the model behaves as predicted by real traffic data.

Details

PDF

Statistics

from
to
Export
Download Full History