| Compared with developed countries,China’s mixed traffic flow consisting of motor vehicles,non-motor vehicles and pedestrians has its own uniqueness,and it is also an important reason for the frequent traffic congestion and accidents in China.Since 2016,shared bicycles have been developing at a rapid rate.The increase in shared bicycles has led to an increase in the proportion of non-motor vehicles in mixed traffic flows,which in turn affects the traffic congestion of urban road network nodes and the probability of traffic accidents.Based on the characteristics of China’s current transportation,this paper takes non-motor vehicles in urban roads as the research object,and analyzes the non-motor vehicle illegal crossing behavior and its influencing factors aturban roadsignalizedintersections.This paper selects 7 signalized intersections in Xi’an,and collects a total of 1 654 cyclist’s endurance time samples through video capture.Kapian-Meier analysis is used to endure all non-motorized vehicles including censored data.The distribution of data is analyzed to explore the distribution law of non-motorized crossing time at signalized intersections.At the same time,based on the complex network theory,the non-motor vehicle network at the intersection is established,and the propagation model of non-motorized illegal street crossing behavior is proposed.Through the simulation analysis of python program,under different network structures and different transmission rates,the behavioral trend of non-motor vehicle violations is obtained,and the two indicators of extroversion and transmission rate are used to characterize the psychological factors of herding.On this basis,combined with other potential influencing factors,the Coxhazardmodel of non-motor vehicle crossing behavior at urban road signalizedintersections is constructed from several aspects: herd mentality,cyclist characteristics,natural environment status,traffic environment status and road environment status.Analysis of non-motorized illegal crossing behavior and influencing factors.The results show that the endurance time of non-motor vehicles is affected by factors such as herd mentality(extroversion and transmission rate),rider gender,cyclist age,non-motor vehicle type,temperature,red light duration,motor vehicle flow,and number of passing lanessignificantly.At the same time,the Binomial Logit model is used to verify the effect of the model.The goodness of fit indicates that the Cox hazardmodel model has a better fitting effect.Finally,the influencing factors such as vehicle flow,red light duration,transmission rate and outwarditude are applied to the model,and a series of standardized strategies to improve the behavior of non-motor vehicles crossing the street are proposed based on the application results.The research in this paper enriches the existing research on the non-motor vehicle illegal crossing behavior at the signalized intersection,and can provide a theoretical basis for the signalized intersection timing signal design,supporting facilities design and traffic management control measures. |
