Non-motorized transport participants(pedestrians,bicycles,electric bikes)are vulnerable to traffic conflicts as vulnerable groups in traffic.pedestrians,vehicles,roads,and the environment will induce conflict between motor vehicles and non-motor vehicles,among which,the right-turn lane in the road factor,as an important traffic infrastructure,plays the function of improving the efficiency of right-turning traffic flow,but it also affects the crossing behavior of non-motorized traffic participants,and different types of right-turn lanes have different impacts on the crossing behavior of non-motorized traffic participants.In order to clarify the changing behavior of non-motorized traffic participants under different types of right-turn lanes,this paper takes two traditional right-turn lanes as the research object,and collects the crossing behavior data of non-motorized traffic participants at six intersections based on the roadside shooting of drones and cameras,of which 1971 people arrived during the red light,and their behaviors can be divided into:waiting in the waiting area,waiting in the overflow waiting area,and running a red light;891 people arrived during the green light period,and their behavior can be divided into:safe passage and risk passage.On this basis,a multinomial and binomial Logit model is established to analyze the influence of right-turn lane type,traffic environment,individual attributes and other factors on the crossing behavior of non-motorized traffic participants,and quantitatively express the influence of each factor with an odds ratio.In addition,in order to further analyze the risky crossing behavior(overflow,red light running,risky traffic)of different types of right-turn lane crossing groups,YOLOv5+Deep SORT was used to extract the trajectory of road users in each intersection,and the risk of risky crossing behavior was analyzed by pass distance(S_P)and post-encroachment time(PET)by safety evaluation index.The results showed that when the right turn lane was on the right side of the road side zone and adjacent to the non-motorized road,the probability of overflowing the waiting area when participants crossed the street was reduced by 71.1%,and the probability of risky passage was reduced by 38.0%.It indicates that the crossing behavior of non-motorized traffic participants under the right turn lane is more standardized.Among the traffic environment factors,the increase in the flow of right-turning traffic will reduce the probability of overflowing the waiting area and running a red light,but it will also increase the probability of risky passage,the increase of the remaining time of the red light will promote the occurrence of the red light behavior,the increase of the remaining time of the green light will reduce the probability of risky passage,and the increase in the length of the crosswalk will lead to an increase in the probability of participants running red lights and risky traffic.Among the personal characteristics,distraction,etc.had a significant impact on street crossing behavior.The probability of running a red light when distracted and accompanied decreased by 73.1%and 56.8%.The probability of cyclists overflowing the waiting area and running a red light increased by 1.782 and 5.948 times,the probability of risky passage decreased by73.8%.The probability of overflowing the waiting area and risk passage when children were walking together decreased by 59.9%and 79.2%,respectively,and the probability of overflowing the waiting area of older people increased.In the risk behavior conflict analysis,the passing distance(S_P)value of the overflow waiting area participant under the type 2 right turn lane was smaller,but the post-encroachment time(PET)value of the red light runner and the risk passer was also smaller.This study provides a reference for the planning and optimization of traditional right-turn lanes and the formulation of traffic management policies. |