An improved algorithm for saccadic intrusion qualification under air traffic control situation

Abstract

Saccadic intrusions are saccadic deviations that interrupt fixations. Saccadic intrusions have the unique characteristics of regular saccades and a small amplitude and round trip, providing a foundation for accurately identifying and quantifying saccadic intrusions. However, currently, rapid and accurate identification algorithms for saccadic intrusions of air traffic controllers are lacking. This study aimed to improve Tokuda's quantification algorithm for saccadic eye movement deviation in terms of data preprocessing, saccadic eye movement identification, and saccade amplitude threshold. The eye movement data were recorded using an eye tracker at a frequency of 100 Hz and obtained from the control simulation experiment. A total of 20 participants with basic control knowledge participated in this study, and finally 54 sets of data were obtained. The analysis of eye movement data showed that the saccadic intrusion quantification value of 36 of 54 sets of data obtained by the improved algorithm was closer to the value output by the manual identification method compared with the original algorithm. The results showed that these improvements had positive effects on the algorithm.