基于眼动指标的飞行员注意状态识别可行性研究  

Feasibility Research on Pilot Attention State Recognition Based on Eye Metrics

作　　者：张益凡 王宇超 张琴喻 葛贤亮 徐杰[1] Zhang Yifan;Wang Yuchao;Zhang Qinyu;Ge Xianliang;Xu Jie(Zhejiang University,Hangzhou 310058,China)

机构地区：[1]浙江大学,浙江杭州310058

出　　处：《航空科学技术》2022年第4期39-46,共8页Aeronautical Science & Technology

基　　金：航空科学基金(20185576005)。

摘　　要：在未来高度自动化的飞行器中,飞行员的作业将以监控任务为主,飞行员注意状态的实时检测和干预技术是保证飞行安全的重要基础性技术。本文采用心理学的缓发持续操作任务(GradCPT)试验范式和眼动追踪技术,探索采用眼动指标来识别飞行员注意状态的可行性。试验招募了30名被试验者,分别完成约30min的GradCPT任务,同时利用眼动仪收集眼动数据。根据被试验者作反应时的特征,被试验者在试验过程中的注意状态被划分为“专注”和“非专注”两种状态。眼动数据显示,瞳孔大小的平均值和变异性、扫视持续时间、扫视峰值速度以及注视持续时间均随注意状态的波动而发生显著的变化。该结果表明,通过眼动追踪技术来识别人的注意状态是可行的。未来的研究可以通过机器学习的方法,以眼动指标作为输入特征,构建针对飞行员的注意力实时监测系统。The main tasks of the pilots in future highly automated aircraft will be tasks monitoring. One of the promising ways to promote performance and safety in such tasks is the development and utilization of real-time pilot attention recognition and intervention technologies in the cockpit. The purpose of the current study is to explore the feasibility of using eye metrics to identify the different states of sustained attention with the Gradual Onset Continuous Performance Task(GradCPT) paradigm. 30participants participated in the study and each completed three eight-minute sessions of GradCPT tasks, while eye metrics are collected with an eye-tracker. "In the zone" and "out of the zone" periods during the task are identified according to the variance time course measure men. Eye-tracking data indicated that the mean and variation of pupil size, saccade duration, saccade peak velocity, and fixation duration measures are sensitive to the "in-the-zone" vs. "the out-of-the-zone" periods. The results suggest that it is feasible to recognize the attention state through eye metrics. Future research can use eye-tracking indicators as input features to build machine learning models of sustained attention for pilots.

关 键 词：人机工效 人机交互界面 航空人因学 持续注意 缓发持续操作任务 

分 类 号：V223.1[航空宇航科学与技术—飞行器设计]