Analysis of Air Traffic Incidents using Event Trees with Fuzzy Probabilities
Jacek Skorupski , Michał Lower , Jan Magott
AbstractThis paper presents a methodology for analysing Air Traffic Incidents based on an estimate of the probability of an Air Traffic Incident developing into an Air Traffic Accident (Air Traffic Incidents are officially defined as events which are dangerous but without a catastrophic impact). A new approach is presented as previous research in this area has focused more on defining accident occurrence probabilities. Access to information on Air Traffic Incidents is restricted and for this reason the methodology developed was based on Fuzzy Set Theory. Analysis was carried out using Event Trees leading from a real Incident to a hypothetical Accident with the probabilities of occurrence of the various scenarios being defined by fuzzy sets. The results of this analysis enable the calculation of the fuzzy probability of the Incident being transformed into an Accident. Performing this analysis, new measuring techniques for the comparison of fuzzy sets were developed and partially verified. The Case Study presented in the paper analyses a Serious Runway Incursion Incident. This Serious Incident is analysed for influencing factors such as: pilot and flight controller skill levels, airport traffic volume, weather conditions, airport procedures and airport geometry. Applying the methods presented in this paper enables an assessment of the effectiveness of Preventive Recommendations of Accident / Incident Investigation Commissions. Additionally, probability estimations can be developed for different incident situations allowing the identification of Security System Weak Points thus enabling appropriate proactive measures to be taken for their elimination.
|Journal series||Fuzzy Sets and Systems, ISSN 0165-0114|
|Publication size in sheets||1.2|
|Keywords in English||air traffic safety, event tree, fuzzy probability, linguistic modelling, serious incident, transportation|
|Score|| = 40.0, 28-11-2017, ArticleFromJournal|
= 40.0, 28-11-2017, ArticleFromJournal
|Publication indicators||: 2016 = 2.718 (2) - 2016=2.783 (5)|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.