K. Zhu, B. Yi, S. Patnaik

Pages: 69-80

Abstract
The empirical analysis method of researching the influencing factors of road traffic accidents has significant importance in deepening the understanding of accident mechanisms, guiding traffic management, and improving traffic safety. In order to solve the problem of low accuracy in traditional methods of analysis, a new empirical analysis method of factors influencing road traffic accidents based on structural equation modeling is proposed. Road traffic accident data is collected, and an improved KNN algorithm is used to repair the collected data. Based on principles such as comprehensiveness, relevance, and representativeness, a preliminary index system for road traffic accident influencing factors is constructed, and grey relational degree is used to select the influencing factor indicators. Based on this, research hypotheses are proposed, and a structural equation model is constructed combining measurement models and structural models. The model is then subjected to procedures such as least squares estimation and correction, with the road traffic accident influencing factor indicators as variables of the structural equation model, to achieve empirical analysis of the structural equation model. The empirical analysis results show that factors such as personnel, vehicles, roads, environment, and management have a direct impact on the probability of road traffic accidents. The average accuracy of the analysis is 98.27%, indicating that it can be widely applied in the field of empirical analysis of road traffic accident influencing factors.
Keywords: structural equation modeling; road traffic accidents; influencing factors; empirical analysis; grey relational degree