A Hybrid Fuzzy-Belief Network (HFBN) for Modelling Aviation Safety Risk Factors

Current analytical risk models of operationally significant events in the aviation domain often lack the ability to produce meaningful and relevant management recommendations since multiple causalities of known and sometimes unknown risk factors need to be analyzed in an integrated systematic fashion. This paper presents a new, combined analytical approach for risk modeling of aviation events. The hybrid approach combines features from fuzzy set theory, with its emphasis on expressing vagueness, with the probabilistic formalism of a Bayesian belief network that captures uncertainty aspects. The resulting hybrid fuzzy-belief network (HFBN) provides an enriched analytical framework with the expressive power to capture the complex realities of aviation events. A wake turbulence encounter scenario is used to validate the proposed model.