Dynamic Bayesian Network for Probabilistic Modeling of Tunnel Excavation Processes

A dynamic Bayesian network (DBN) model for probabilistic assessment of tunnel construction performance is introduced. It facilitates the quantification of uncertainties in the construction process and of the risk from extraordinary events that cause severe delays and damages. Stochastic dependencies resulting from the influence of human factors and other external factors are addressed in the model. An efficient algorithm for evaluating the DBN model is presented, which is a modification of the so-called Frontier algorithm. The proposed model and algorithm are applied to an illustrative case study, the excavation of a road tunnel by means of the New Austrian Tunneling Method.