Research on Construction Tower Crane Safety Management Based on FDA-DBN

Based on Forecast-Decision-Action theory, this study analyzes the causes of tower crane accidents from the perspective of safety information transmission, dividing the safety risk factors of tower cranes into three levels involving six main stakeholders and a total of 17 influencing factors. Furthermore, the FDA risk system is transformed into a Dynamic-Bayesian-Network (DBN). By applying fuzzy set theory to calculate the probability of risk factors, the DBN structure is used for parameter learning to analyze the likelihood of tower crane accidents, identify key influencing factors, and trace the paths of influence. The research indicates that current tower crane construction should focus on workers’ behavior, and various departments need to strengthen the division of responsibilities in tower crane supervision to reduce the probability of accidents.