Research on Influencing Factors of Urban Road Traffic Casualties through Support Vector Machine

Abstract

Urban road traffic safety has always been vital in transportation research. This paper analyzed the factors influencing the degree of traffic accident casualties on Jinan Jingshi Road and its branch roads, taking them as the study area for urban road traffic safety problems. Additionally, it used the application of Particle Swarm Optimization (PSO), a Support Vector Machine (SVM) model, and a recursive feature elimination (RFE) to rank the contribution degree of the influencing factors. The results showed that driving on rainy days has a high probability of casualties, while the type of collision was a minimum influence factor. Additionally, on rainy days, cars were accident-prone road vehicles, and 8:00–12:00 and 18:00–22:00 were accident-prone periods. Based on the results, preventive measures were further put forward regarding the driver, road drainage capacity, policy management, and autopilot technology. This study aimed to guide urban traffic safety planning and provide a basis for developing traffic safety measures.