飞行模拟器环境下飞行员视觉特征  

Visual features of pilot based on Markov chain under flight simulator

作　　者：孙瑞山[1] 汪苧[1] 

机构地区：[1]中国民航大学民航安全科学研究所,天津300300

出　　处：《北京航空航天大学学报》2013年第7期897-901,共5页Journal of Beijing University of Aeronautics and Astronautics

基　　金：国家自然科学基金资助项目(60979009);国家"973"计划资助项目(2010CB734105);中央高校基本科研业务专项资金资助项目(ZXH2010A001)

摘　　要：为了研究飞行员在各个阶段的视觉特征,构建了以固定式飞行模拟器为基础、采用先进的非接触式眼动仪FaceLAB4.0采集数据的实验测试平台.根据飞行员获取信息的来源将视觉区域划分为舱内仪表盘、前窗舱外景和其他3个部分,应用马尔科夫链的数学方法分析被试者的视觉注视数据在各视觉区域内的状态转移概率.研究发现,被试在观察舱内仪表盘和前窗舱外景区域时,无法在短时间内获得全部信息,需要长时间注视才能完成信息获取.该方法可用于分析飞行员在操作中的视觉注意力和注意力转移模式.In order to study the visual features of pilot in difference flight phases, an integrated flight sim- ulator was used to built an experimental platform, and a kind of non-contact eye tracker named FaceLAB4.0 was used to track the data of the tester. The tester was asked to complete the takeoff task. Based on the visual information sources, the visual interested areas of pilot were divided into three parts including instrument pan- el, outside and others. The data of the tester were recorded and the method of Markov chain was used to ana- lyze state transition probability of the visual gaze data in various visual areas. The study indicates that the pilot is generally need to gaze for a long time, because there are more information clues in cockpit instrument panel and outside areas. This method could be used to analyze pilot＇s visual attention and its transfer pattern in oper- ation.

关 键 词：眼动跟踪 马尔科夫链 视觉特征 状态转移 

分 类 号：X949[环境科学与工程—安全科学]