Studies of vehicle lane-changing to avoid pedestrians with cellular automata

This paper presents studies of interactions between vehicles and crossing pedestrians. A cellular automata system model of the traffic is developed, which includes a number of subsystem models such as the single-lane vehicle model, pedestrian model, interaction model and lane-changing model. The random street crossings of pedestrians are modeled as a Poisson process. The drivers of the passing vehicles are assumed to follow a safety-rule in order not to hit the pedestrians. The results of both single and multiple car simulations are presented. We have found that in general, the traffic can benefit from vehicle lane-changing to avoid road-crossing pedestrians. The traffic flow and average vehicle speed can be increased, which leads to higher traffic efficiency. The interactions between vehicles and pedestrians are reduced, which results in shorter vehicle decelerating time due to pedestrians and less switches of the driving mode, thus leads to the better energy economy. The traffic safety can be improved in the perspective of both vehicles and pedestrians. Finally, pedestrians can cross road faster. The negative effect of lane-changing is that pedestrians have to stay longer between the lanes in the crossing.