Bayesian Network Risk Method for Complex System Engineering Operations

This study proposed a risk assessment method based on Bayesian networks for complex system engineering with multiple types of work, strong concealment, and multiple uncertain factors. Starting from the analysis of the engineering operation process, various influencing factors that may bring risks were identified, and a Bayesian network risk assessment model was established. To address the problem of lacking input data for multi-state root nodes, a triangular fuzzy function was applied to determine the probability value, thus to address the difficulty in obtaining conditional probability values for multi-state nodes, a knowledge matrix was constructed and DS evidence theory is applied to fuse and process expert information. Simulation analysis shows that Bayesian network causal reasoning can determine engineering operation risk indicators, diagnostic reasoning can determine key influencing factors and reasonable preventive improvement measures can be formulated based on this.