当前位置:主页 > 社科论文 > 心理论文 >

模拟管制中的眼动数据处理算法及应用研究

发布时间:2018-06-15 08:41

  本文选题:模拟管制 + 数据处理算法 ; 参考:《中国民航大学》2017年硕士论文


【摘要】:近年来,我国民航取得累累硕果,安全水平处在世界前列。在民航运输快速发展的同时,空管系统面临的问题越来越严峻,空管运行保障能力严重不足与民航快速发展之间的矛盾日益突出。空管“错、忘、漏”导致的不安全事件时有发生。对管制员工作进行实时监控可以有效减少管制员的工作失误,提高航空飞行效率和安全系数。管制员工作所需信息几乎全部来自视觉,通过眼动行为对管制员工作进行研究可以掌握管制员的工作特点、行为决策等。对于眼动数据的处理是眼动研究的基础与关键。本文在综合本实验室研究及国内外对于眼动数据处理研究的基础上,提出改进的眼动数据处理算法。首先对原始眼动数据进行预处理,删除空白数据、消除噪声数据、转换数据格式;将眼动坐标数据、飞机坐标数据进行坐标转换;将预处理后的数据利用离散傅里叶变换得到最佳距离值,进而得到最佳注视序列;用最佳注视序列指导眼动数据的聚类。在算法基础上,设计了“模拟管制注意力分配可视化系统”;系统实现了眼动数据、飞机数据、视频数据的导入,关键参数设置,时间窗分析,分析结果导出等功能。利用算法及系统开展应用研究。对管制员与管制学员注视过程差异及注视转移进行了研究,发现:专家组形成的注视点数比新手组多且各个被试之间注视点数比较均衡,离散程度小,整体表现优于新手组,关注过程注视的差异跟管制员的管制经验和工作能力密切相关;管制员有优于管制学员的管制策略和管制习惯,大部分被试都是按照“环形”模式进行管制,管制学员没有固定的转移模式及其目标,注意力分布相对离散,按照星型模式进行注意转移的居多。对模拟管制中管制员遗忘规律进行了研究,发现:被试对管制场景信息的回忆中,对于航班呼号遗忘较为严重,对于速度、高度、尾流等级的遗忘率在50%左右,对于跑道号则不会遗忘;总结了被试的遗忘规律:任务型遗忘、空闲型遗忘、等待型遗忘、穿越型遗忘、其它型遗忘。对模拟管制中管制员扫视模式进行了研究,提出了扫视类型的定义及扫视类型的识别方法;通过扫视类型的识别方法对模拟管制实验中10名被试的扫视模式进行了挖掘;通过挖掘总结了四种扫视模式:三角型、线型、混合型、其它型。
[Abstract]:In recent years, China's civil aviation has made numerous achievements, safety level is in the forefront of the world. With the rapid development of civil aviation transportation, the air traffic control system is facing more and more serious problems, and the contradiction between the serious shortage of air traffic control operation support capacity and the rapid development of civil aviation is becoming increasingly prominent. Air traffic control "wrong, forget, leak" lead to unsafe incidents occur from time to time. Real-time monitoring of controller work can effectively reduce the controller's working errors and improve the efficiency and safety factor of aeronautical flight. Almost all the information needed by the controller comes from vision. Through the study of eye movement behavior, we can master the working characteristics and behavior decision of the controller. The processing of eye movement data is the foundation and key of eye movement research. In this paper, an improved algorithm of eye movement data processing is put forward on the basis of synthesizing the research of our laboratory and the research of eye movement data processing at home and abroad. Firstly, the original eye movement data is preprocessed, the blank data is deleted, the noise data is eliminated, the data format is transformed, and the eye movement coordinate data and the plane coordinate data are transformed. The preprocessed data are obtained by using discrete Fourier transform (DFT) to obtain the best gaze sequence, and the best gaze sequence is used to guide the clustering of eye movement data. On the basis of the algorithm, the visual system of simulated control attention allocation is designed, which realizes the functions of eye movement data, aircraft data, video data import, key parameter setting, time window analysis, analysis result derivation and so on. Use algorithm and system to carry out application research. The differences in fixation process and fixation transfer between controllers and cadets were studied. It was found that the fixation points formed by the expert group were more than those of the novice group, and the fixation points among the subjects were more balanced, the dispersion was small, and the overall performance was better than that of the novice group. The differences in attention to process fixation were closely related to the controller's control experience and ability to work; controllers had better control strategies and habits than the trainees, and most of the participants were regulated in a "circular" mode. There is no fixed transfer mode and its target, and the attention distribution is relatively discrete, most of them are based on star pattern. In this paper, the law of controller forgetting in simulated control is studied. It is found that the forgetting rate of flight call sign is more serious than that of flight call sign, and the forgetting rate of wake grade is about 50%. The forgetting rules of the subjects were summarized: task-based forgetting, idle forgetting, waiting forgetting, traversing forgetting and other forgetfulness. In this paper, the controller scan pattern in analog control is studied, and the definition of scan type and the recognition method of scan type are put forward, and the scanning pattern of 10 subjects in simulated control experiment is excavated through the recognition method of scan type. Through mining, four kinds of scanning patterns are summarized: triangle type, linear type, mixed type, and others.
【学位授予单位】:中国民航大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:B842

【参考文献】

相关期刊论文 前10条

1 袁枭;王炎鑫;宋绪政;;Java语言的特点与C++语言的比较研究[J];科技创新与应用;2016年28期

2 靳慧斌;王丹;王松涛;;民航管制员的注视及遗忘特征研究[J];科学技术与工程;2016年16期

3 田少煦;申品品;郭昱竹;;基于眼动跟踪技术的色彩情感研究[J];现代传播(中国传媒大学学报);2015年06期

4 马锦飞;常若松;高远;;瞳孔直径大小检测驾驶员疲劳的实证效度分析[J];辽宁师范大学学报(社会科学版);2014年01期

5 国威;潘晓东;邓其;喻泽文;;驾驶员隧道段夜间行车动视点特征模糊聚类评价[J];公路工程;2013年06期

6 张杰;;基于眼动仪的驾驶员视点分布特性研究[J];湖南交通科技;2012年04期

7 葛岩;郑雪梅;;眼睛是心灵的窗口——基于眼动追踪技术的网页可用性测试[J];现代传播(中国传媒大学学报);2012年02期

8 冯成志;贾凤芹;;平滑追随眼动的神经机制综述[J];应用心理学;2008年01期

9 张卫东;骆慈孟;;试论专职空中交通管制教员在管制员培训工作中的必要性[J];空中交通管理;2002年06期

10 蔡勉,王磊,肖国镇;离散傅里叶变换在序列线性复杂度快速算法中的应用[J];西安电子科技大学学报;2001年05期

相关博士学位论文 前3条

1 关伟;驾驶员对交通标志的视觉信息认知过程实验研究[D];北京工业大学;2014年

2 平源;基于支持向量机的聚类及文本分类研究[D];北京邮电大学;2012年

3 袁伟;城市道路环境中汽车驾驶员动态视觉特性试验研究[D];长安大学;2008年

相关硕士学位论文 前8条

1 王丹;雷达管制中的视觉搜索模式研究[D];中国民航大学;2016年

2 董斌;扇区交通运行特征与管制员眼动行为相互关系研究[D];南京航空航天大学;2016年

3 蔡亚敏;模拟管制过程中的动态目标视觉注视转移特征研究[D];中国民航大学;2015年

4 单慧;基于JAVA的订餐管理系统的设计与实现[D];大连理工大学;2015年

5 张琳;眼睛注视知觉及其对面孔再认和表情判断等社会认知过程的影响[D];西南大学;2013年

6 李士刚;空管“错忘漏”事件发生机理与预警方法研究[D];中国民用航空飞行学院;2012年

7 李晓菲;数据预处理算法的研究与应用[D];西南交通大学;2006年

8 刘鑫朝;聚类算法的研究及其应用[D];长沙理工大学;2006年



本文编号:2021434

资料下载
论文发表

本文链接:https://www.wllwen.com/shekelunwen/xinlixingwei/2021434.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户c9505***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com