Risk Identification Of Civil Aviation Unsafe Events Based On Correlation Of Classification Variables

In recent years,the civil aviation transportation industry has developed faster and faster,but there have been many major civil aviation accidents every year.The overall aviation safety situation is not optimistic.Therefore,it is more and more important to strengthen the investigation of aviation safety and control the occurrence of civil aviation accidents.This paper mainly studies the civil aviation safety issues from two aspects:Firstly,the characteristics,data structure and related relationship of civil aviation unsafe events are studied.Due to the complex data structure of civil aviation insecure events,in order to explore the relationship between various categorical variables,this paper introduces a concept tree model and uses multi-dimensional association rules to explore the potential relationship between categorical variables of unsafe events.At the same time,in order to study and analyze the highly relevant but infrequently related association attributes,this paper sets two different support thresholds and combines the same trust threshold to conduct data mining analysis.Second,the risk identification of civil aviation unsafe events.In this paper,the wavelet decomposition and reconstruction method is used to denoise the risk value of the bird strike event,and then the risk identification is carried out for the high-risk time period in a specific period.The Bow-tie model is established to analyze the risk of civil aviation unsafe events.As an example of insecure events,specific examples are analyzed from the airport and route.The Poisson fitting method for the occurrence of civil aviation unsafe events is established to estimate the probability of future unsafe events.Finally,the association rules and The mapping is combined to identify the causal factors of civil aviation unsafe events and to monitor the key factors causing the key factors.Introduce the cause of accident and risk transfer theory,and classify the causes of unsafe events into four dimensions:"human-machine-loop-tube".Using the mining of association rules,mapping the causal factors and finding the causative factors The key factors between the two are to strengthen the monitoring of the risk factors and achieve control over civil aviation unsafe events.