Civil Aviation Safety Risk Assessment for Rare Events

摘要

Civil aviation safety accidents are rare, and are typical small-probability risk accidents. However, no accidents do not mean that there are no safety hazards. Analysis of such small-probability and minimum probability of safety risks, for the accurate assessment of civil aviation safety risk levels, Targeted management strategies are important. This paper proposes a method for mapping small-probability events to the ANN using fault tree mapping to overcome the limitations of the fault tree model in dynamic risk analysis. This paper takes the civil aviation maintenance system as an example to analyze the safety of the accident caused by the engine, establish a fault tree and map it one by one, and use the artificial neural network for multiple training. Using the example, the results are compared with the FT results. It is found that the matching rate between ANN and FT results is high, indicating that the model of ANN with FT mapping is a technique with reasonable risk assessment. The method proposed in this paper can also be applied to other high-reliability comprehensive systems for example nuclear oriented plant systems and power structures.