MODELLING THE PEDESTRIAN CROSSING SPEED USING FUZZY LOGIC

Abstract

Despite all the measures taken, highway traffic accidents have an ever increasing trend in Turkey. In 2017, 1.202.726 traffic accidents occurred in Turkey, resulting in 7.427 deaths and 300.383 injuries. The pedestrian faults in this accidents have a share of 8.3%. The majority of these accidents occur when the pedestrians share the road with the motor vehicles. Pedestrian crossings are designed or the motor vehicle traffic is stopped using traffic lights for safe crossing of pedestrians. However, in most cases, the duration of the traffic lights is determined only by taking into account the motor vehicle traffic parameters, ignoring the crossing speeds of the pedestrians. In this study, pedestrian crossings were examined at different intersections on the Anatolian side of Istanbul, which was selected as pilot region. Using the data obtained, pedestrian crossing speeds are modelled using fuzzy logic, taking into consideration the individual characteristics, weather conditions, traffic conditions and time conditions. Using the developed model, the time that pedestrians need to cross the street can be determined in advance. In this way, more appropriate pedestrian crossing designs can be made or a real-time traffic control system which takes the pedestrian traffic into account realistically can be designed.