A Novel Method for Analysis of Key Causes of Aviation Insecurity Events Based on Mutual Information

Based on mutual information, this paper proposes a method of aviation safety dimension reduction, which is used to construct the causation model of aviation accidents under uncertain environment. First, based on the Bow-tie model, a comprehensive m comprehensively considering many factors such as the cause of the accident, possible consequences, and control measures. Aiming at the problem of high-dimensional modeling of causal factors, a mutual information measure was proposed to measure the causal strength between causal factors and aviation safety, and screen out causal variables that have a greater impact on aviation safety. For mutual information needed to solve the problem of multidimensional variables cause of joint integration it is proposed based on sampling in conjunction with low deterministic kernel density estimation method for solving joint integration, convergence and high accuracy. Taking aviation safety as an example, the correctness and scientificity of the method in this paper are verified.