基于多传感器数据的航空管制员情境意识的计算

航空管制人员保持良好的情境意识是确保空中交通以安全、高效方式运行的必要条件。如果在管制员工作过程中, 能准确监测其情境意识状态, 并在情境意识丧失前给予及时告警, 那么将会极大地提高管制绩效和减少空中两机相撞的概率。目前, 国外评价管制员情境意识的方法虽然不少, 但这些方法基本都是诊断性质和回溯性质的, 无法对管制员的情境意识进行监测或预警。管制员的情境意识受个体差异、任务等多种因素的影响, 同时, 其也可以由某些指标来表征, 如某些作业绩效指标和生理指标。为了能够实现对航空管制员情境意识的准确监控及告警, 拟通过雷达管制模拟实验、心理测量、专家评定等方法采集多传感器数据, 分别建立基于表征指标和基于影响变量的情境意识计算模型, 并比较其预测性能。该研究结果可为管制员情境意识的实时监测提供支持。

A Computational Approach to Air Traffic Controller Situation Awareness Based on Multi-sensor Data

Good situation awareness (SA) is highly important for air traffic controllers (ATCos) to maintain the safe, orderly and expeditious flow of air traffic. If ATCos’ situation awareness can be accurately monitored and alerted before it has been lost, it will be highly possible to improve ATCos’ performance and prevent accidents. At present, many ways exist to evaluate ATCos’ situation awareness, but they are basically diagnostic and retrospective methods, early warning and forecasting cannot be achieved. ATCos’ situation awareness could be affected by many factors, but it can be demonstrated by some performance and physiological indicators. In order to improve the prediction accuracy of SA, multi-sensor data which are relevant to ATCos’ situation awareness will be collected by radar control simulation experiment, psychological measurement, and expert ratings. The ultimate aim of the research is to establish SA computational models by two different approaches, which address demonstration indicators and contributing variables of the above multi-sensor data respectively. An evaluation of the two prediction Models will also be made. The research results are expected to provide supports for monitoring ATCos’ situation awareness.