A LANE-CHANGING BML MODEL CONSIDERING THE INFLUENCE OF BOTH LANE INFORMATION AND TURN SIGNALS

The lane-changing behaviors of vehicles are frequently adopted by drivers in order to get better driving conditions and turn signal plays an essential role in guiding vehicles' lane-changing behaviors. In this paper, based on the BML (Biham-Middleton-Levine model ) model, we proposed a lane- changing cellular automata model which takes the influence of both lane information and turn signals into account. When a vehicle cannot move forward, the driver will judge whether the vehicle meets the lane-changing conditions or not. If the driver can change his/her lane, then the lane-changing probability is calculated according to the lane information (such as average velocities and densities of those vehicles in his own and target lanes) and the state of turn signals. Finally, the driver can determine whether he/she changes lane or not by the corresponding probability. Numerical simulations were carried out to investigate the effect of lane-changing behaviors on the phase transition between the free flow phase and the global jamming phase. Two kinds of BML models were studied, one with traffic lights and the other without traffic lights. Numerical results show that the critical density of the BML model without traffic lights increases considerably due to the introduction of lane-changing rules. At a smaller scale, the critical density approximates that of the BML model with traffic light control. The effect of lane change is significant on traffic dynamics. Furthermore, a new coexisting phase of both the free flow phase and the local jamming phase was found. The underlying generation and evolution mechanism of the coexisting phase is discussed in detail. It is shown that the local congestions will result in the global congestion under higher densities. However, the lane-changing rules does not have distinct effect on the critical density of the BML model with traffic lights. But the region of phase transition becomes narrower. It indicates that lane-changing behaviors can lead to variations of local traffic features and take less effect on the global features of traffic system.