Comparative Study On Crossing Behaviors Of Pedestrians And Nonmotorized Vehicles At Signalized Intersections

Compared with the developed districts such as Europe and the United States, traffic violations by pedestrians and nonmotorized vehicles at urban road is more prevalent, and the traffic accident rate has been higher in China. The red light running behavior of pedestrians and bike riders is one of the most typical and multiple rule violation behaviors in the cities. Analysis, comparison, and exploration the inherent regularity of all kinds of subtle crossing behavio is the basis of proposing targeted management and control strategy for pedestrians and nonmotorized vehicles. Therefore, it has vast theoretical and practical importance for enhancing the traffic safety of pedestrians and nonmotorized vehicles. Accordingly, based on the field observation data, using log linear model and logistic regression model, this paper compare all kinds of subtle crossing behavior before and during the crossing process between pedestrians and nonmotorized vehicles, such as temporal and spatial related behavior, visual search behavior, and further analysis the effect of subtle crossing behavior to red light running behavior. Specifically, the contents of this dissertation are as follows:(1) Subtle crossing behaviors before crossing among different modes are analyzed and compared. Using log linear model to control gender and age, results indicate that there are significant differences among three modes on the behaviors related to speed, temporal and spatial compliance, visual research, distractors before crossing. Compared with pedestrians, nonmotorized vehicle cyclists are more likely to accelerate or decelerate when arriving intersections, more disperse in the lateral waiting position, and have higher rate waiting on the reasonable position of longitudinal waiting position than pedestrians. Compared with pedestrians and bicyclists, e-bike cyclists have higher rate in looking moving vehicles and lower rate in looking signal light before crossing. In addition, pedestrians have higher rate in chatting than cyclists during the process of waiting.(2) Subtle crossing behaviors during crossing among different modes are analyzed and compared. Using log linear model to control gender and age, results indicate that there are significant differences among three modes on the behaviors related to temporal and spatial compliance, visual research, distractors, speed, cross path during crossing. Pedestrians and e-bike cyclists are much more likely to violate in the phase of turn-left of far road; the proportion that pedestrians walk in the zebra crosswalk is much higher than the proportion that cyclists rider in the reasonable district during crossing; the proportion of pedestrians and bicyclists to look dangerous vehicles is significantly higher than that of e-bike cyclists; the proportion of bicyclists who use phone is significantly higher than that of pedestrians and e-bike cyclists; Compared with pedestrians, cyclists are more likely to accelerate or decelerate during crossing.(3) Analysis of main risking crossing behavior and its influencing factors. Logistic regression models were used to establish the model of immediate crossing behavior for all arrivals and the model of red light running behavior for waiters only. The former model results indicate that people who is middle aged, arrives at phase of turn-left of far road, or when less people are waiting, or when more people are breaking rule, he will more likely to crossing immediately after arriving. The later model results indicate that people who is e-bike cyclist, waiting in the unreasonable position, waiting position has changed, looking vehicles or not looking signal light during waiting process, or when there are less people getting to wait during waiting process, or when there are more people breaking rule, he will more likely to running red light.