This article presents the results of studies of the interrelation between the parameters that characterize the psychophysiological state of the pilot and the quality index of piloting.To estimate the pilot’s state,si...This article presents the results of studies of the interrelation between the parameters that characterize the psychophysiological state of the pilot and the quality index of piloting.To estimate the pilot’s state,signals of various natures were used:deviations of the aircraft control stick in the pitch and roll channels,electroencephalograms,determination of emotional states,and analysis of blinking parameters from video images of the face.As quality index of piloting,vertical and horizontal deviations from the glide path line during landing approaches on the flight simulator were considered.Signal spectral densities,frequency coherence functions,and principal component analysis were used to process experimental data,and convolutional deep learning neural networks were used to analyze video images.As a result of the research,stable correlations were found between the index of piloting accuracy and such characteristics of the operator as the frequency coherence functions of the control signals,the ratio of the first and second principal components of the electroencephalogram signals,the number of emotional states,and the number of blinks.This newly found correlation between data of different nature is to be useful for pilot interface design.展开更多
基金supported by Russian Foundation for Basic Research(RFBR),project 20-08-0049.
文摘This article presents the results of studies of the interrelation between the parameters that characterize the psychophysiological state of the pilot and the quality index of piloting.To estimate the pilot’s state,signals of various natures were used:deviations of the aircraft control stick in the pitch and roll channels,electroencephalograms,determination of emotional states,and analysis of blinking parameters from video images of the face.As quality index of piloting,vertical and horizontal deviations from the glide path line during landing approaches on the flight simulator were considered.Signal spectral densities,frequency coherence functions,and principal component analysis were used to process experimental data,and convolutional deep learning neural networks were used to analyze video images.As a result of the research,stable correlations were found between the index of piloting accuracy and such characteristics of the operator as the frequency coherence functions of the control signals,the ratio of the first and second principal components of the electroencephalogram signals,the number of emotional states,and the number of blinks.This newly found correlation between data of different nature is to be useful for pilot interface design.
