Air traffic controller fatigue detection method, device and system based on cluster analysis

Abstract

The present invention provides a method, device and system for air traffic controller fatigue detection, said method comprising: step 1, receiving control seat video monitoring data and land-air call data obtained by a real-time audio and video full data acquisition device; The video monitoring data of the control seats and the ground-air call data are detected and analyzed to obtain the controller fatigue index detection data; step 3, refer to the preset controller fatigue index detection data samples, and use the cluster analysis method to construct the controller's work status file In step 4, the controller fatigue index detection data in step 2 is classified and comprehensively detected, and if the similarity exceeds the threshold range, an alarm is triggered. The fatigue detection method can be applied in the field of air traffic control, and has strong practicability and operability.

Description

technical field [0001] The invention provides an air traffic controller fatigue detection method, device and system integrating face and voice, belonging to the technical field of image processing and pattern recognition. technical background [0002] At present, the relevant research on controller fatigue is mainly carried out from two aspects, namely, controller fatigue detection based on facial recognition, and controller fatigue detection based on workload evaluation. The former is based on the analysis of the facial features of the controller, and obtains the data of three fatigue detection indicators, such as PERCLOS (percentage of eyelid closure) value, average eye closure time, and yawn frequency, and matches the three indicators of controller fatigue judgment. The corresponding threshold is used to detect controller fatigue. For example, the application number is CN104112335A, which provides a driver fatigue detection method based on multi-information fusion, which ...

AI Technical Summary

Problems solved by technology

[0003] The existing research on controller fatigue detection mainly has the following deficiencies: (1) In terms of research methods, there are more qualitative research and less quantitative research, resulting in insufficient objectivity; (2) In terms of detection indicators, the index dimension is relatively single and comprehensive (3) Existing studies use the same index detection standard for all controllers, ignoring the individual differences in fatigue indicators, and the detection results are not accurate enough; (4) In terms of applicability, it is oriented to the actual situation of air traffic control units. There are few engineering applications and it is difficult to realize

Due to the above deficiencies, the current research on controller fatigue detection at home and abroad is lacking in objectivity, comprehensiveness, comprehensiveness, accuracy and operability.

Images