Research On Flight Situational Awareness Based On Fatigue Monitoring

In modern warfare,battlefield situations are becoming increasingly multi-dimensional,where pilots’ primary work has gradually shifted from flight control to situational awareness and they are suffering from a surge of situational awareness pressure.Multi-dimensional situational information and the ever-accelerating tempo of the battlefield are increasing pilots’ mental load and are challenging their cognitive limitations;under such circumstances,pilots tend to get cognitively overloaded and may make mistakes in perception and decision making.With the development of artificial intelligence,machine intelligence could help human intelligence with perception and decision making in complex contexts.The intelligent integration of human and machine requires an intelligent task allocation aligning with pilots’ changing states;therefore,comprehensively monitoring and understanding the capability states of pilots are the basis of realizing intelligent cockpits.From the perceptive of pilots’ fatigue states,this paper combines theoretical analysis and empirical study to analyze and test the monitoring technique of flight fatigue and the relationship between situational awareness and fatigue level.The primary research and innovations of this paper include:(1)Propose a multi-modal flight fatigue measurement combining electrocardiogram and eye metrics.Based on the analysis of the characteristics of cockpit environment and physiological measurements,the study of methods to extract and analyze physiological metrics,and a combination of empirical simulated flight experiments,research proposes a flight fatigue measurement based on electrocardiogram and eye metrics.This measurement integrates multiple modalities to enhance the reliability of measuring process and could be utilized to achieve a lightweight and non-invasive detection of flight fatigue.(2)Propose an effective technique for automated flight fatigue classification.Through the study of fatigue classification techniques and a combination of application and verification with the experimental dataset,an automated classification technique for flight fatigue is proposed that combines Toeplitz inverse covariance clustering to identify fatigue ground truth and supervised learning algorithms to build the automated classification models.This technique avoids the subjective bias of traditional methods during ground truth identification and could realize an accurate and automated classification of fatigue states.(3)Verify the correlation hypothesis between flight situational awareness and fatigue states.Based on the theoretical study of situational awareness and its assessment methods,the research proposes a hierarchical evaluation method that can divide different levels of situational awareness and makes the theoretical assumptions about the correlation between different situational awareness levels and fatigue states.Finally,through experimental study,research implemented the hierarchical evaluation and verify the theoretical hypothesis;the different correlation between perception level,comprehension level and projection level and fatigue states were found.Current research results provide technical support for pilot state monitoring and can provide design basis for the realization of intelligent human-machine collaboration and adaptive system functions based on flight fatigue monitoring.