Risk of vehicle rear end collision in function of traffic flow

Many virtual reality experiments pointed out that the risk accepted by drivers crucially depends on his workload and traffic interferences. Even if rear end collisions are the most frequent road accidents on highway, traffic flow has not been adequately taken into account in the project. This is the consequence of the traditional approach that neglects human factors and driver’s behaviour. This study has analysed the risk of rear end collision using a driving simulator. The main objective is to verify the variability of probability and severity of accident for different traffic flows. Particularly five traffic scenarios representing different levels of service have been developed for simulations. Twenty subjects have been selected for the experiments. During simulation cinematic and dynamic parameters are recorded. Moreover time/space distances between driver’s car and other vehicles have been recorded. Output post-processing has provided interesting results for the evaluation of the variability of subjective risk accepted by the driver in function of the level of service. The objective risk, computed as the product of probability and severity, has been computed for each scenario as well.