Abstract:

To effectively control the risks associated with bridge jacking renovation construction, a safety risk analysis method of bridge jacking renovation construction based on DBN was proposed. By considering the diverse risk factors and their dynamic changes over time in bridge jacking renovation construction, a risk index system was established from four aspects: human, material, environmental, and management. A DBN model was constructed by incorporating the time dimension, with network node probabilities quantified using fuzzy theory and expert scoring. The conditional probabilities were refined through the Leaky Noisy-or Gate extension model. Bidirectional inference was performed using the established DBN model to dynamically analyze the safety risks of bridge jacking renovation construction. The results indicate that environmental factors are the primary risk contributors to bridge jacking renovation construction, while material factors have a lesser impact. Weather, jack failure, on-site construction management, and substructure instability are identified as the most sensitive risk factors, warranting focused prevention measures during bridge jacking renovation construction.

Key words: dynamic Bayesian network (DBN), bridge jacking renovation construction, construction safety, risk analysis, conditional probability, risk factor