Skip to Main Content
The nature of flight risk assessment has always been a problem of uncertainty. In order to quantify the risk somehow is important. Based on the previous related studies of airlines flight safety assessment a new flight risk assessment index system was built in this paper. Bayesian belief networks model with diagnostic and predictive inference was proposed to evaluate flight accident probability on the knowledge and experience of experts. Then by the use of two-grade fuzzy comprehensive evaluation with employing centralization statistical method, the accident consequence severity level could be obtained. The application through a practical example shows that the risk assessment can not only avoid the subjectivity of the experts but also deals with the uncertain information effectively. The assessment results can objectively and comprehensively exhibit the risk of flight for airlines, and the method turns out to be a feasible way to the flight risk assessment.