Analysis of Influencing Factors of Accident Morphology Based on Logistic Regression

In order to explore the influencing factors of urban inter-vehicle accident patterns, the traffic accident data of a city are used in this paper to study the influencing factors of different accident forms through the disordered multivariate logistic regression model. Also, the probability of accidents in various forms is predicted, and the model test is significant. According to the results of the study, the influencing factors of different accident forms are also different. However, pedestrian crossings, flexible guardrails, and visibility are generally the protective factors for traffic accidents, while machine-non-mixed road and lighting, etc. are generally the risk factors for accidents.