Influence of driver’s yielding behavior on pedestrian-vehicle conflicts at a two-lane roundabout using fuzzy cellular automata

The roundabouts are widely used in China, some of which have central islands as scenic spots. The crosswalks connecting to the central islands, normally full of pedestrians, have negative impact on roundabout capability and pedestrian safety. Therefore, this study proposes a fuzzy cellular automata (FCA) model to explore the safety and efficiency impacts of pedestrian-vehicle conflicts at a two-lane roundabout. To reason the decision-making process of individual drivers before crosswalks, membership functions in the fuzzy inference system were calibrated with field data conducted in Changsha, China. Using specific indicators of efficiency and safety performance, it was shown that circulating vehicles can move smoothly in low traffic flow, but the roundabout system is prone to the traffic congestion if traffic flow reaches to a certain level. Also, the high yielding rate of drivers has a negative impact on the traffic efficiency but can improve pedestrian safety. Furthermore, a pedestrian restriction measure was deduced for the roundabout crosswalk from the FCA model and national guideline of setting traffic lights.