Quantitative Analysis Of CFIT Events Based On Bayesian Network

As a resource on which safety management activities depend,aviation safety information is an important channel for obtaining and identifying safety risks.In-depth analysis and quantification can provide a data basis for safety risk management and accident prevention.At present,the amount of aviation safety information collected has reached a certain scale,but it is still difficult to give full play to the value of information in terms of analysis and utilization.Information quantification still remains at the statistical level of quantity,due to insufficient access to the value of the information contained,it is not possible to give full play to the early warning function that the information should have.Therefore,it is very necessary to conduct in-depth quantitative analysis of aviation safety information,and provide targeted support for risk control and safety assurance.This paper takes the CFIT(Controlled Flight into Terrain)event as the research sample.First,the Bayesian network method was used to establish a CFIT event quantitative model including the event consequence layer,the aircraft status layer and the inducing factor layer;then a method for calculating the information entropy of aviation safety events based on the event consequences was proposed and established,the event model experiment analyzed event information entropy,and quantified the amount of event information from the dimension of event consequences;The CFIT events were collected by Civil Aviation of China from 2017 to 2019 for quantification,and cleaned the event samples as sample one.Then the event information entropy of sample 1 was calculated,and the influence intensity of each of the event consequence layer,the aircraft status layer,and the inducing factor layer was calculated,the importance of inducing factors was measured and the probability of occurrence of aircraft status and event consequences was got.In view of the insufficient quantification of accident event information,to continue to expand the sample of this type of event,the CFIT events were collected by the aviation safety network(ASN)from 2000 to 2019.Finally,the quantitative results of the total sample were obtained.Compared the quantitative results,safety recommendations were gained to prevent such events.The following conclusions were drawn through research:1、In the CFIT event model,the impact of event information entropy has the relationship that accident is greater than incident,and incident is greater than general event;2、The samples of CFIT events in China’s civil aviation industry were quantitatively analyzed,it is concluded that the risk prevention and control of CFIT events focus on the top three inducing factors are "flight support","altitude control" and "situational awareness" with their corresponding quantitative values,and the causal event chain that predicts the greatest probability of a CFIT event with quantitative value is that the flight support factor causes the height to be too low,which in turn leads to the general event.3 、 After expanding the sample of aviation safety net events,the total sample was analyzed comprehensively and quantitatively,finally it is concluded that the risk prevention and control of CFIT events focus on the top three inducing factors are "altitude control","flight guarantee" and "situational awareness" with their corresponding quantitative values,and the causal event chain that predicts the greatest probability of a CFIT event is that the altitude control factor causes the height to be too low,which in turn leads to the accident.Based on the quantitative results,suggestions for the prevention and control of such incident risks are strengthening the flight altitude control of the crew,strengthening the crew’s situational awareness,and attaching importance to the construction of the flight support environment.At the same time,the quantification process of this type of event provides a quantitative idea for other unsafe event information,improves the problem of insufficient acquisition of aviation safety information value,and provides strong support for risk control and safety assurance.