Human Factor Analysis Model of Civil Aviation Incidents Based on Bayesian Network

Abstract

To generate a model which can provides detailed data analysis supportive in aviation incident analysis, a human factor analysis model based on Bayesian network theory is established. This model is a Bayesian network which uses three layers nodes to represent causality between human factors and incidents. The specific impact degree of human factors on aviation incidents is represented by conditional probability parameters of the model. The model structure, constructed by combing hill-climbing search method with CH score function, coincides with the actual data. This model is useful in aviation incident analyses and deductions.