Dynamic evaluation of road and bridge engineering construction safety risk based on Fuzzy Dynamic Bayesian Network research.

The article presents a dynamic safety risk evaluation model for road and bridge engineering construction based on Fuzzy Dynamic Bayesian Network (FDBN), integrating Fuzzy Set Theory (FST) with an improved similar aggregation method to reduce subjectivity in expert assessments. Using the Hebi City Provincial Highway 304 reconstruction project as a case study, the model dynamically predicts safety risk probabilities over time, identifies key risk factors—such as lifting and hoisting accidents, high-altitude falls, collapses, mechanical equipment accidents, and food poisoning—and aligns well with actual construction site conditions. The approach combines expert opinions, accident reports, and probabilistic reasoning to support decision-making and optimize risk management strategies. The study suggests future work to enhance coupling analysis of risk factors and incorporate probabilistic loss considerations.