Evaluation of construction safety resilience of water conservancy projects based on D-S evidence theory and intuitionistic fuzzy Bayesian networks

To improve the construction safety level of water conservancy projects and carry out research on construction safety resilience, this paper proposes a safety resilience evaluation method integrating intuitionistic fuzzy set theory and Bayesian network model. From the three dimensions of absorptive capacity, recovery capacity and adaptive capacity, the influencing factors of construction safety resilience of water conservancy projects are identified. The interpretive structural model is adopted to analyze the hierarchical relationship among resilience influencing factors and construct the topological framework of Bayesian network. The D-S evidence theory is used to transform experts’ intuitive fuzzy linguistic evaluation results into probability parameters of network nodes. The topological framework of Bayesian network is applied to carry out forward reasoning calculation of resilience, the dominant path chain of resilience is analyzed combined with backward reasoning, and the key influencing factors of resilience are identified through sensitivity analysis. Taking the Comprehensive Regulation Project of the Lower Reaches and Tailwaters of the Ganjiang and Fuhe Rivers as a case study, the construction safety resilience is evaluated. The results show that the probability of high safety resilience of the project construction system reaches 71.74%, reflecting an overall good level of construction safety resilience. The dominant resilience path chain is safety organization and management structure-emergency response plan-emergency disposal capacity-resilience, and the safety monitoring system together with emergency response plan serve as the key influencing factors.