Association mining of factors affecting truck traffic accidents based on improved Apriori algorithm

To explore the underlying mechanics of truck traffic accidents, identify key factors, and implement effective safety management strategies, a method based on an enhanced Apriori algorithm for determining critical causes of such accidents is proposed. This study analyzed 1839 truck accident records from a mountainous freight highway in Guangdong Province, China. Using the improved Apriori algorithm, 571 association rules were discovered among all factors and specific dimensions (time, road conditions). Findings indicate significant support for rules such as inadequate following distance, improper operation, involvement of 1-2 vehicles, clear weather, interactions between trucks and cars, slopes exceeding 2%, and curve radii under 700 meters—all suggesting drivers' unsafe behaviors leading to accidents. Moreover, association rule mining revealed elevated lift values (>1.4) for accidents involving injuries between 1-3 am, and for minor accidents during the 8-9 am and 5-6 pm periods, highlighting peak accident times. Additionally, significant accidents were linked to downhill sections with 3%-4% slopes and radii exceeding 1000 meters, emphasizing their high-risk nature. These insights into truck accident correlations provide technical support for highway improvement and safe operations in freight transport.