Research On The Characteristics Of Pedestrian-vehicle Conflicts And Safety Evaluation Of Advance Right-turn Lane

With the accelerated process of urbanization,there has been a sharp increase in residents’ travel demands,leading to increasingly severe traffic pressure on urban road networks.Traffic congestion and driving safety at intersections have become important issues faced by countries worldwide.One commonly used channelization method for large intersections is the advance right-turn lane.Compared with the general right-turn method,the advance right-turn lane widens the entrance on the right side and sets up a turning bend to divert right-turning vehicles into the left flow of straight-through vehicles,effectively reducing delays for right-turning vehicles and improving traffic efficiency.However,the design of the diversion island results in pedestrians crossing the street twice,and the lack of signal control mode makes it difficult to avoid the risk of conflicts between pedestrians and vehicles in the curved area.Therefore,studying the characteristics of conflicts between pedestrians and vehicles in advance right-turn lanes and conducting safety evaluations are of great engineering value.This paper conducted on-site investigations and collected traffic flow data from advance right-turn lanes,statistically analyzed the conflict characteristics between pedestrians and motor vehicles,established a multivariate ordinal regression model for conflict severity to reveal the influence factors of pedestrian-vehicle conflicts,and proposed a pedestrian crossing safety evaluation model.The main contents of this paper include the following aspects:(1)The study aims to analyze the mechanism of conflicts between pedestrians and motor vehicles at the right-turn lane in advance.The path data of pedestrians and motor vehicles in the study area were collected to examine safety issues related to mixed traffic flow.Tracker and Premiere Pro software were utilized to process the basic video investigation materials,extracting features of pedestrian crossing and pedestrian-vehicle movement paths.Homogeneous coordinate transformation was applied to convert the path data from image coordinates to world coordinates,laying the foundation for the study of the characteristics of pedestrian crossing and pedestrian-vehicle conflict at the rightturn lane in advance.(2)The study examines the distribution characteristics of pedestrian crossing paths and defines types of spatial paths.It proposes a Pedestrian Path Deviation(PPD)index to quantitatively analyze the distribution characteristics of pedestrian paths in curved areas.Furthermore,it conducts cross-analysis and modeling research on individual characteristics and crossing behavior of pedestrians,exploring the impact of individual differences on pedestrian crossing behavior patterns.(3)The severity of conflicts between pedestrians and vehicles is calculated from both temporal and spatial dimensions using indicators such as time to collision,postencroachment time,and deceleration to safety time.K-means clustering is utilized to classify the severity levels of conflicts into three categories: potential,minor,and severe.Four independent variables are selected from the perspectives of pedestrians,vehicles,roads,and the environment,with the addition of the PPD index as an independent variable.A multiple ordered regression model for conflict severity is established,revealing the impacts of different factors on conflicts between pedestrians and vehicles.(4)We constructed a safety evaluation index model for conflicts between vehicles and pedestrians in advance right-turn lanes.The model includes primary evaluation indexes of pedestrians,vehicles,and road environment,and secondary evaluation indexes such as traffic volume,crossing speed,channel design,and pedestrian crossing behavior.We determined the weights of the evaluation indexes by using analytic hierarchy process,grey correlation degree,and objective-subjective combination of information entropy.The cloud model was utilized to convert quantitative and qualitative language,completing the safety risk evaluation of advance right-turn lanes.An example has verified the effectiveness of the model,and measures to improve traffic safety in advance right-turn lanes were suggested from three aspects: pedestrians,motorized vehicles,and road facilities.