Assessment of UAV's Operator Cognitive State Based on Behavior Signals

基于行为学的无人机操作员认知状态评估

School of Electronics and Information, Northwestern Polytechnical University, Xi’an 710072, China

Received: 20 May 2017

Abstract

For exploring the relationship between the mental or cognitive state and metric of vigilance test for unmanned aerial vehicle (UAV), a vigilance state evaluation method and sphere of application based on behavior signals is established. A classical vigilance test avoiding to crash is set. During the experiments, the subjective ratings as well as behavior signals (Response Time, Lapse) are recording. The dynamic changing of behavior signals is analyzed using statistical analysis. The results demonstrate that compared with continuous PVT test, the subject's mental workload in rest PVT test decreases dramatically. Compared other metrics, the speed of response time can reflect the dynamic changing of subject's mental state. The metric of Q-50 has a strong robustness for outlier of subject. Considering that the metrics have strong correlation with operator's cognitive state, they can effectively analyze the different workload.

摘要

为探索无人机操作员的精神或认知状态与警觉度测试度量准则的关系，研究不同准则的应用环境，建立基于行为信号的警觉度评价方法。设置了不同条件下经典的警觉度防撞击实验，记录实验过程中被试的主观评分和响应时间、错误率等行为学信号。通过统计学方法分析实验过程中行为参数的动态变化。结果表明：相比于连续性任务，间歇性任务下被试的主观精神负载显著降低；相比于其他度量准则，响应速率更能反映被试的精神状态的动态变化；相比于平均响应时间，Q-50对极值具有更高的鲁棒性；当被试警觉度状态较高或较低时，Q-10和Q-90的变化范围较小或不变。