Evaluation on Severity of Pedestrian-Vehicle Traffic Conflicts at Signalized Intersection Based on SOM

Abstract:
In order to evaluate the traffic safety of pedestrians crossing the intersections, this paper regards the traffic conflicts between pedestrians and vehicles at signalized intersections as the research object, extracts 403 four - dimensional microscopic data based on trajectory tracking technology, and conducts the Self-Organizing Map (SOM) algorithm to determine the severity level of traffic conflicts. Based on the results of SOM clustering analysis, the safety evaluation of intersections is carried out and the reliability of the classification result is tested perfectly compared with the practical situations. This research can provide a new perspective for traditional safety evaluation methods.