| Aviation safety information is an important basis for the identification of safety risks,the control of safety risks,and the improvement of aviation safety.In addition to the accident and incident information,employees’ voluntary report information is an important information source to improve the level of aviation system operation.Mining the hidden safety risks in the voluntary reports,carding safety risks which can cause serious consequences,establish the semantic networks contain causes and consequences,which can provide the basic data for the further analysis of safety risks.Taking 200 flight crew fatigue reports information from ASRS as a sample,the flight fatigue time,flight phases,causes,consequences were analyzed by using the semantic network analysis method.The flight fatigue semantic network has been built according to the process which concludes key information extraction,semantic network framework,node and nodes relationship determent.The relationship between nodes has been confirmed by using the grey correlation analysis method.Prevention and control measures for the flight fatigue semantic network nodes have been analyzed by using Bow-Tie model,which can refined the semantic network for the description of causes and consequences of fatigue events.Taking a fatigue report in 2015 as an example to analyze,the empirical analysis results show that: firstly,the flight fatigue semantic network can easily draw the causes and consequences of fatigue event.Secondly,according to the results of the node correlation analysis,other probable consequences that may happen can also be found,and measures can be found in the Bow-Tie model to prevent these consequences.Thirdly,the new causes and consequences of the information can be added into the current semantic network and make its contents more rich and perfect,and provide support reference to the control of fatigue risk.In summary,the semantic network analysis method can analyze knowledge and their relationship,in addition to the analysis of flight fatigue event reports,it also provides a reference for the analysis of other same voluntary reports.Currently the semantic network in voluntary report analysis is still in the artificial analysis stage,because the characteristics of the report which has no fixed order requirements,in the future,analysis tools can be used in the report analysis to improve the analysis efficiency. |
