Correlation analysis of factors affecting self-driving takeover accidents based on Apriori algorithm
TANG Wenzhi
Abstract：The purpose of this paper is to explore the risk factors and their interdependence of autonomous vehicle driving takeover collision accidents.The collision data of 659 autonomous vehicle released by the California Motor Vehicle Administration from January 2018 to June 2024 were collected.By using association rule mining method,identify the set of risk factors that frequently appear together in autonomous vehicle driving takeover collisions.Research has found that the risk of accidents is closely related to factors such as road design(such as the presence of a median strip and no roadside parking),weather conditions(such as humidity and dryness),and vehicle movement status(such as driving straight).For example,there is a strong bidirectional correlation between road design with a median strip and no roadside parking and driving takeover.In humid weather,one-way lane accidents are more likely to occur.The research results can provide reference for policy measures and engineering countermeasures to improve road safety and efficiency of autonomous vehicle,so as to improve the overall safety and reliability of road traffic.
Keywords：accident analysisassociation rulesrisk factorsautonomous vehicledriving takeover
Publication Date：2025-10-25
Pages：4( 23-26 )