Research on risk assessment of deep foundation pit construction based on Dynamic Bayesian Network

A risk assessment method based on the Dynamic Bayesian Network for deep foundation pit construction is proposed to aim at the complexity, uncertainty, and dynamics during the deep foundation pit construction process. Firstly, the work decomposition structure-risk decomposition structure is used to extract and combine the risk factors from five aspects, namely, personnel, materials, machinery, technology, and environment, and then collated into a risk evaluation index system concerning the relevant standards, specifications, and consulting experts' opinions. Secondly, the DEMATEL method is used to determine the causal relationship between these risk factors, which is then mapped into the dynamic Bayesian network graphical structure. Finally, the fuzzy set theory and the improved similar aggregation method are used to determine the prior probabilities and conditional probabilities of the network nodes, and the conditional probabilities are corrected by using the Leaky Noisy-or Gate extension model; the transfer probabilities between different time slices are determined based on the Markov Chain assumptions, to construct a comprehensive and scientific Dynamic Bayesian Network model. To provide new ideas for analyzing and effectively controlling the safety risks of deep foundation pit construction.