基于眼动数据的管制员注意力特征评价

doi:10.16265/j.cnki.issn1003-3033.2023.02.2785

摘要/Abstract

摘要：

为提高管制员培训的针对性和有效性,提升管制过程的安全性,提出使用注意力稳定性、注意力分配与转移和注意力广度3个指标,作为注意力特征的评价指标;设计并开展管制过程模拟管制试验,采集分析36名成熟管制员的眼动数据;提出选取兴趣区访问次数方差(VVC)、马尔可夫转移概率(MTP)和特情反应时间(RT)3个参数,将其作为刻画注意力特征3个评价指标对应的量化参数值,再使用K-Means聚类分析法对管制员的注意力特征进行分类,得到管制员注意力特征的评价值的结果分类区间,并对比评价结果与教员打分结果。结果表明:根据注意力特征,将管制员分成4类,一类为最优,四类为最差。优秀管制员具有注意力稳定性好、注意力分配合理、注意广度大和对特情敏感的特点,其量化参数值包括:VVC为13.48~95.46,MTP为0.03~0.22,RT为3~31 s。

关键词: 管制员, 眼动数据, 注意力特征, K-Means聚类, 马尔可夫转移概率(MTP)

Abstract:

In order to improve the relevance and effectiveness of air traffic controllers' training and enhance the safety of air traffic control process, the attention stability, attention allocation and transfer and attention breadth were proposed as the evaluation indexes of attention characteristics firstly. Secondly a simulated control experiment was designed to collect and analyze eye-movement data of thirty-six mature controllers. Thirdly three parameters of the variance visit counts(VVC) of area of interests, MTP and emergency situation reaction time(RT) were selected as the quantitative parameter values corresponding to the three evaluation metrics that portray attentional characteristics. Finally, K-Means clustering analysis was used to classify the attentional characteristics of controllers, and the resultant classification intervals of the evaluation values of the attentional characteristics of controllers were obtained. The evaluation results were compared with the expert scoring results. The results show that there are obvious differences in the attention characteristics and evaluation value of controllers with different categories. Excellent controllers have the characteristics of good stability, reasonable distribution of attention, wide attention span and sensitivity to special situations. The quantization parameter values are the VVC to the area of interest is 13.48-95.46, the interval of MTP is 0.03-0.22, and the special situation RT interval is 3-31s.

Key words: air traffic controller, eye movement data, attention characteristics, K-Means clustering, Markov transition probabilities(MTP)

王莉莉, 许凌鹏. 基于眼动数据的管制员注意力特征评价[J]. 中国安全科学学报, 2023, 33(2): 217-224.

WANG Lili, XU Lingpeng. Evaluation of air traffic controller's attention characteristics based on eye movement data[J]. China Safety Science Journal, 2023, 33(2): 217-224.

图/表 12

图1

图2

表1

表2

表3

图3

表4

图4

表5

表6

表7

图5

参考文献 19

[1]	MCCARLEY J S, WICKENS C D. Applied attention theory[M]. Boca Raton: CRC Press, 2008:42.
[2]	CHRISTINA K, LAURA H I, CRAIG H, et al. Using eye-tracking to investigate strategy and performance of expert and novice control room operators[J]. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2014, 58(1):1667-1671. doi: 10.1177/1541931214581348
[3]	MANDAL S, KANG Z. Using eye movement data visualization to enhance training of air traffic controllers: a dynamic network approach[J]. Journal of Eye Movement Research, 2018, 11(4): DOI:10.16910/jemr.11.4.1. doi: 10.16910/jemr.11.4.1
[4]	邓文颖. 影响管制员注意行为的因素和注意力分配的模式[J]. 空中交通管理, 2000(6):53-55.
	DENG Wenying. The factors influencing ATC controller's focusing attention and the model of attention distribution[J]. Air Traffic Management, 2000(6):53-55.
[5]	王超, 于超博, 王敏. 基于注意力分配的管制员调配飞行冲突认知过程研究[J]. 安全与环境学报, 2016, 16(4):205-209.
	WANG Chao, YU Chaobo, WANG Min. On the cognitive process of the air traffic controllers for the growing fight conflicts based on the attention distribution[J]. Journal of Safety and Environment, 2016, 16(4): 205-209.
[6]	李敬强, 王蓓, 赵宁, 等. 基于K-Means聚类的管制员注意品质特征研究[J]. 中国安全科学学报, 2017, 27(6):13-18. doi: 10.16265/j.cnki.issn1003-3033.2017.06.003
	LI Jingqiang, WANG Bei, ZHAO Ning, et al. On characteristics of attention quality of controller based on K-Means clustering[J]. China Safety Science Journal, 2017, 27(6): 13-18. doi: 10.16265/j.cnki.issn1003-3033.2017.06.003
[7]	李敬强, 李康, 王勇, 等. 管制员信息加工水平鉴别技术及影响机制[J]. 中国安全科学学报, 2018, 28(6):25-30. doi: 10.16265/j.cnki.issn1003-3033.2018.06.005
	LI Jingqiang, LI Kang, WANG Yong, et al. Research on identification technology and influence mechanism of information processing levels of controllers[J]. China Safety Science Journal, 2018, 28(6):25-30. doi: 10.16265/j.cnki.issn1003-3033.2018.06.005
[8]	王茜. 管制负荷与眼动特征的关系研究[D]. 天津: 中国民航大学, 2019.
	WANG Qian. Study on the relationship between controller's workload and eye movement characteristics[D]. Tianjin: Civil Aviation University of China, 2019.
[9]	BAO Haifng, FANG Weining, GUO Beiyuan, et al. Supervisors' visual attention allocation modeling using hybrid entropy[J]. Entropy, 2019, 21(4): DOI:10.3390/e21040393. doi: 10.3390/e21040393
[10]	张美玉. 基于管制员眼动行为的注意力分配研究[D]. 南京: 南京航空航天大学, 2019.
	ZHANG Meiyu. Air traffic controller's attention allocation based on eye movements[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2019.
[11]	王莉莉, 朱敏. 基于事故树的空中交通管制员情景意识分析[J]. 安全与环境学报, 2021, 21(1):249-256.
	WANG Lili, ZHU Min. On the situational awareness of the air flight controller based on the fault tree analysis[J]. Journal of Safety and Environment, 2021, 21(1): 249-256.
[12]	DUCHOWSKI A T. Eye tracking methodology: theory and practice[M]. Cham: Springer International Publishing AG, 2017:11-13.
[13]	王莉莉, 陈凤兰. 基于脑电的管制员认知行为与疲劳的关系研究[J]. 中国安全科学学报, 2018, 28(7):1-6. doi: 10.16265/j.cnki.issn1003-3033.2018.07.001
	WANG Lili, CHEN Fenglan. Study on relationship between controllers' cognitive behavior and fatigue based on EEG[J]. China Safety Science Journal, 2018, 28(7): 1-6. doi: 10.16265/j.cnki.issn1003-3033.2018.07.001
[14]	中国民用航空局. 雷达管制基础模拟机培训大纲[Z], 2012-06-26.
[15]	邓肯, 王梓坤. 马尔可夫过程论基础[M]. 哈尔滨: 哈尔滨工业大学出版社, 2015:20-32.
[16]	TORBJÖRN F, NILS P G. Fixation patterns of learner drivers with and without cerebral palsy (CP) when driving in real traffic environments[J]. Transportation Research Part F: Psychology and Behaviour, 2001, 4(3):171-185. doi: 10.1016/S1369-8478(01)00021-3
[17]	杨勇. 管制员情景意识的量化研究[D]. 天津: 中国民航大学, 2020.
	YANG Yong. Quantitative research on controllers' situational awareness[D]. Tianjin: Civil Aviation University of China, 2020.
[18]	唐飞, 王云刚, 杜炳成, 等. 基于优化马尔可夫模型的煤矿事故死亡人数预测[J]. 中国安全科学学报, 2022, 32(4): 122-128. doi: 10.16265/j.cnki.issn1003-3033.2022.04.018
	TANG Fei, WANG Yungang, DU Bingcheng, et al. Prediction of death toll in coal mine accidents based on optimized grey Markov model[J]. China Safety Science Journal, 2022, 32(4): 122-128. doi: 10.16265/j.cnki.issn1003-3033.2022.04.018
[19]	杨俊闯, 赵超. K-Means聚类算法研究综述[J]. 计算机工程与应用, 2019, 55(23):7-14. doi: 10.3778/j.issn.1002-8331.1908-0347
	YANG Junchuang, ZHAO Chao. Survey on K-Means clustering algorithm[J]. Computer Engineering and Applications, 2019, 55(23): 7-14. doi: 10.3778/j.issn.1002-8331.1908-0347

评价指标	被试1	被试2	…	被试35	被试36
VVC	16.53	19.41	…	41.67	54.88
MTP	0.086	0.189	…	0.389	0.333
RT	50	15	…	20	11

评价指标	被试1	被试2	…	被试35	被试36
VVC	0.04	0.06	…	0.18	0.26
MTP	0.19	0.42	…	0.89	0.76
RT	0.76	0.19	…	0.27	0.13

指标		VVC	MTP	RT
VVC	皮尔逊相关性	1	0.021	-0.216
	双尾Sig.	—	0.905	0.207
	个案数	36	36	36
MTP	皮尔逊相关性	0.021	1	0.08
	双尾Sig.	0.905	—	0.642
	个案数	36	36	36
RT	皮尔逊相关性	-0.216	0.08	1
	双尾Sig.	0.207	0.642	—
	个案数	36	36	36

指标	聚类中心
指标	1	2	3	4
VVC	0.57	0.38	0.16	0.57
MTP	0.14	0.38	0.64	0.14
RT	0.15	0.41	0.21	0.15

指标		管制员分类	教员打分
管制员分类	皮尔逊相关性	1	-0.706^**
	双尾Sig.	—	0.000
	个案数	36	36
教员打分成绩	皮尔逊相关性	-0.706^**	1
	双尾Sig.	0.000	—
	个案数	36	36