DANGEROUS SITUATIONS IN TWO-LANE TRAFFIC FLOW MODELS

Abstract

This paper investigates the probability of car accidents (PCA) in two-lane traffic flow models. We introduce new conditions for the occurrence of dangerous situations (DS) caused by an unexpected lane changing vehicles. Two different lane changing rules are considered, say symmetric and asymmetric. For the symmetric rules, we investigate the influence of the Nagel–Schreckenberg parameters such as the maximal speed, the randomization probability, …, on the PCA when vehicle moves forward or changes lanes. It is found that the forward PCA is as likely as that in one-lane traffic model. As regards to lane changing, the properties of the PCA are qualitatively different from those in one-lane traffic. For the asymmetric rules, we investigate the effect of the slack parameter Δ, introduced to adjust the inversion point of lane-usage, on the PCA. Contrarily to one-lane traffic, the forward PCA in the right lane exhibits two maximums for some range of Δ; the first one is located at low density and the second at high density. The lane changing PCA from right to left is found to decrease with increase of Δ. However, no DS exist when vehicles change from left to right.