In view of the pedestrian space violation in an advance right-turn lane, the pedestrian crossing paths are divided by collecting the temporal and spatial information of pedestrians and motor vehicles, and the characteristics of different pedestrian crossing behaviors are studied. Combined with the time and speed indicators of conflict severity, the K-means method is used to divide the level of conflict severity. A multivariate ordered logistic regression model of the severity of pedestrian-vehicle conflict was constructed to quantify the effects of different factors on the severity of the pedestrian-vehicle conflict. The study of 1388 pedestrians and the resulting pedestrian-vehicle conflicts found that the type of spatial violation has a significant impact on pedestrian crossing behavior and safety. The average crossing speed and acceleration variation values of spatially violated pedestrians were significantly higher than those of other pedestrians; there is a significant increase in the severity of pedestrian-vehicle conflicts in areas close to the oncoming traffic; the average percentage of pedestrian-vehicle conflicts due to spatial violations increased by 12%, and the percentage of serious conflicts due to each type of spatial violation increased from 18% to 87%, 74%, 30%, and 63%, respectively, compared with those of non-violated pedestrians. In addition, the decrease in the number of lanes and the increase in speed and vehicle reach all lead to an increase in the severity of pedestrian-vehicle conflicts. The results of the study will help traffic authorities to take measures to ensure pedestrian crossing safety.
Keywords: advance right-turn lane; conflict severity; multivariate ordered logistic model; spatial crossing violation; traffic safety.