ABSTRACT
Undercrossing shield tunnels often cause damage to aboveground structures, which are high-risk projects. Reasonable risk assessment is crucial for construction decision-makers and safety managers. In this paper, the risk of building damage caused by an undercrossing Beijing shield metro is assessed using the Bayesian network. The noisy-or model and the Noisy-Max model are used to establish the Bayesian network model. The model is applied to predict the risk probability of building damage induced by the undercrossing shield metro by forwarding reasoning. The results are prepared with the statistical risk probability derived by building settlement monitoring value, which verifies the effectiveness of the model. The key risk factors affecting building safety were diagnosed and analyzed, which can provide a reference for project risk management.
