A Realistic Two-Lanes Traffic Simulation Model Based on Cellular Automata

Lane-changing models are an important component of microscopic traffic simulation. In this paper an asymmetric two-lanes cellular automata (CA) model for traffic flow is presented. The aim is to reproduce the actual usage of two-lanes highway space and to make the lane change modeling more in line with reality, when different types of vehicles are considered. For this purpose, the model determines both the utility of a given lane change and the risk associated with it, by means of a set of lane changing rules in terms of accelerations and safe braking distances. Simulation results on a system with periodic conditions and two types of vehicles with different length and different limited velocities are realized. In contrast to the most CA model existing in the literature, it is shown that the traffic flow is not dominated by the introduction of large vehicles with small maximum velocity, when the change of lane is not prohibited for them, and the lane usage inversion observed in the reality is still reproduced. Moreover, the model avoids, at least partially, multiple lane-changes of vehicles in consecutive time-steps. The model preserves the computational simplicity of CA traffic models, that makes real time implementation feasible.