Modelling Drivers' Behaviour as a Crash Risk Reduction Process

Seyyed Mohammad Sadat Hoseini

Manouchehr Vaziri

Abstract

The evermore widespread use of microscopic traffic simulationin the analysis of road systems has refocused attention onsub models, including car-following and lane-changing models.In this research a microscopic model is developed whichcombines car-following and lane-changing models and describesdriver behaviour as a crash risk reduction process ofdrivers. This model has been simulated by a cellular automatasimulator and compared with the real data. It has been shownthat there is no reason to consider the model invalid for drivers'behaviour in the basic segments of freeways in Iran, duringnot-congested conditions. Considering that uncertainty of positionof vehicles is caused by their acceleration or deceleration, aprobability function is calibrated for calculating the presenceprobability of vehicles in their feasible cells. Multiplying thepresence probability and impact of crash, crash risk of cells iscalculated. As an application of the model, it has been shownthat when difference between vehicles brake deceleration increases,the total crash risk increases.