基于SSENet的飞行员脑力疲劳评估方法

doi:10.19665/j.issn1001-2400.20250204

摘要/Abstract

摘要：

着舰任务具备时间紧迫、操作过程复杂等特点,准确评估飞行员脑力疲劳对提高着舰安全至关重要。针对着舰任务中飞行员脑力疲劳评估问题,开展不同难度模拟着舰实验,采集9名被试人员在为期9天实验中的脑电信号,构建基于SSENet的着舰场景下跨被试脑力疲劳评估模型。针对脑电信号空间特征和跨被试训练方法,在模型中设计了SEConv模块以捕捉脑电信号中的空间信息耦合与通道特征信息。结果表明: 在不同难度的着舰任务中,被试的脑力疲劳程度存在显著性差异(p<0.001),模型在五折交叉验证中取得了最高95.55%的分类准确率,平均分类准确率为93.00%,消融实验验证了各模块有效性,相较于经典脑电信号训练模型EEGNet,分类准确率提升了约4%。SSENet在跨被试脑力疲劳评估任务上取得良好效果,有望对提高着舰安全方向的研究提供新的策略。

关键词: 脑电信号, 脑力疲劳, 飞行员, 着舰, 深度学习

Abstract:

The landing task is characterized by time pressure and a complex operational process,making it crucial to accurately assess pilots' mental fatigue to enhance landing safety.To address the issue of evaluating pilots' mental fatigue during landing tasks,a series of simulated landing experiments of varying difficulties are conducted,with EEG signals from nine participants over a nine-day period collected.A cross-subject mental fatigue assessment model based on SSENet is developed for the landing scenario.To capture spatial information coupling and channel feature information in the EEG signals,an SEConv module is designed within the model,targeting the spatial characteristics of the EEG signals and cross-subject training methods.The results show significant differences in the level of mental fatigue among participants during the various difficulty levels of landing tasks(p<0.001).The model achieves a maximum classification accuracy of 95.55% during five-fold cross-validation,with an average classification accuracy of 93.00%.Ablation experiments verify the effectiveness of each module,with a classification accuracy improvement of approximately 4% compared to the classical EEG signal training model EEGNet.The SSENet demonstrates promising results in the cross-subject mental fatigue assessment task,offering new strategies for enhancing the research on landing safety.

Key words: electroencephalography, mental fatigue, landing, pilot, deep learning

中图分类号:

金恒, 孙有朝, 曾一宁, 刘威成, 郭媛媛. 基于SSENet的飞行员脑力疲劳评估方法[J]. 西安电子科技大学学报, 2025, 52(2): 33-46.

JIN Heng, SUN Yuochao, ZENG Yining, LIU Weicheng, GUO Yuanyuan. Pilot mental fatigue assessment method based on the SSENet[J]. Journal of Xidian University, 2025, 52(2): 33-46.

图/表 17

图1

表1

图2

表2

表3

图3

图4

图5

图6

图7

图8

表4

表5

表6

表7

图9

图10

参考文献 37

[1]	LEE S, KIM J K. Factors Contributing to the Risk of Airline Pilot Fatigue[J]. Journal of Air Transport Management, 2018, 67:197-207.
[2]	HILDITCH C J, GREGORY K B, ARSINTESCU L, et al. Perspectives on Fatigue in Short-Haul Flight Operations from US Pilots:A Focus Group Study[J]. Transport Policy, 2023, 136:11-20.
[3]	LIU B, XIN X, JI M, et al. Can Family-Work Conflict Influence Safety Behavior in Airline Pilots? The Mediating Role of Fatigue and the Moderating Role of Extraversion[J]. Safety Science, 2023,160:106061.
[4]	VAN CUTSEM J, MARCORA S, DE PAUW K, et al. The Effects of Mental Fatigue on Physical Performance:A Systematic Review[J]. Sports Medicine, 2017, 47:1569-1588.
[5]	刘晓雨, 孙立国, 谭文倩, 等. 基于相似构型决策的舰载机驾驶员建模与评估[J]. 航空学报, 2023, 44(4):113-128.
	LIU Xiaoyu, SUN Liguo, TAN Wenqian, et al. Pilot Modeling and Evaluation of Carrier-Based Aircraft Based on Similar Configuration Decision-Making[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(4):113-128.
[6]	王秋富, 石治国, 张倬, 等. 舰载机着舰引导中鲁棒单目视觉相对位姿测量[J]. 航空学报, 2024, 45(23):330309-1-16.
	WANG Qiufu, SHI Zhiguo, ZHANG Zhuo, et al. Robust Monocular Vision-Based Relative Pose Measurement in Carrier-Based Aircraft Landing Guidance[J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(23):330309-1-16.
[7]	王永庆. 固定翼舰载战斗机关键技术与未来发展[J]. 航空学报, 2021, 42(8):21-34.
	WANG Yongqing. Key Technologies and Future Developments of Fixed-Wing Carrier-Based Fighter Jets[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(8):21-34.
[8]	ZENG H, ZHANG J, ZAKARIA W, et al. InstanceEasyTL:An Improved Transfer-Learning Method for EEG-Based Cross-Subject Fatigue Detection[J]. Sensors, 2020, 20(24):7251.
[9]	SHANG L, SI H, WANG H, et al. Research on Fatigue Detection of Flight Trainees Based on Face EMF Feature Model Combination with PSO-CNN Algorithm[J]. Scientific Reports, 2024, 14(1):20641. doi: 10.1038/s41598-024-71192-x pmid: 39232069
[10]	张丞, 何坚, 张岩, 等. 基于脑电与眨眼频率的可穿戴疲劳驾驶检测系统[J]. 计算机工程, 2017, 43(2):293-298. doi: 10.3969/j.issn.1000-3428.2017.02.049
	ZHANG Cheng, HE Jian, ZHANG Yan, et al. Wearable Fatigue Driving Detection System Based on EEG and Blink Frequency[J]. Computer Engineering, 2017, 43(2):293-298. doi: 10.3969/j.issn.1000-3428.2017.02.049
[11]	ADIN R M, CEREN A N, SALCI Y, et al., Dimensionality,Psychometric Properties,and Population-Based Norms of the Turkish Version of the Chalder Fatigue Scale Among Adults[J]. Health and Quality of Life Outcomes, 2022, 20(1):161.
[12]	AHLSTRÖM C, ANUND A. Development of Sleepiness in Professional Truck Drivers:Real-Road Testing for Driver Drowsiness and Attention Warning(DDAW) System Evaluation[J]. Journal of Sleep Research, 2024:e14259.
[13]	SU A T, XAVIER G, KUAN J W. The Measurement of Mental Fatigue Following an Overnight On-Call Duty Among Doctors Using Electroencephalogram[J]. PLoS ONE, 2023, 18(7):e0287999.
[14]	DU G, LONG S, LI C, et al. A Product Fuzzy Convolutional Network for Detecting Driving Fatigue[J]. IEEE Transactions on Cybernetics, 2022, 53(7):4175-4188.
[15]	张荣, 梁馨月. 航空人员疲劳检测方法研究[J]. 计算机工程与应用, 2024, 60(20):84-95. doi: 10.3778/j.issn.1002-8331.2403-0135
	ZHANG Rong, LIANG Xinyue. Research of Fatigue Detection Methods for Aviation Personnel[J]. Computer Engineering and Applications, 2024, 60(20):84-95. doi: 10.3778/j.issn.1002-8331.2403-0135
[16]	谷学静, 刘佳, 郭宇承, 等. 采用多尺度多路混合注意力机制的脑电情绪识别方法[J]. 计算机工程与应用, 2024, 60(19):130-138. doi: 10.3778/j.issn.1002-8331.2309-0201
	GU Xuejing, LIU Jia, GUO Yucheng, et al. EEG Emotion Recognition Method Using Multi-Scale Multi-Path Hybrid Attention Mechanism[J]. Computer Engineering and Applications, 2024, 60(19):130-138. doi: 10.3778/j.issn.1002-8331.2309-0201
[17]	YUAN D, YUE J, XIONG X, et al. A Regression Method for EEG-Based Cross-Dataset Fatigue Detection[J]. Frontiers in Physiology, 2023,14:1196919.
[18]	WU X, YANG J, SHAO Y, et al. Mental Fatigue Assessment by an Arbitrary Channel EEG Based on Morphological Features and LSTM-CNN[J]. Computers in Biology and Medicine, 2023,167:107652.
[19]	赵朔, 奇格奇, 李培豪, 等. 基于脑电通道注意力机制的驾驶行为识别研究[J]. 交通运输系统工程与信息, 2024, 24(4):283-291.
	ZHAO Shuo, QI Geqi, LI Peihao, et al. Research on Driving Behavior Recognition Based on EEG Channel Attention Mechanism[J]. Journal of Transportation System Engineering and Information, 2024, 24(4):283-291.
[20]	邓浩伟, 侯月皎, 张朝月, 等. 基于级联森林和多模态融合的脑力疲劳识别算法[J]. 北京航空航天大学学报, 2025, 51(2):584-593.
	DENG Haowei, HOU Yuejiao, ZHANG Zhaoyue, et al. Brain Fatigue Recognition Algorithm Based on Cascade Forest and Multimodal Fusion[J]. Journal of Beijing University of Aeronautics and Astronautics, 2025, 51(2):584-593.
[21]	冯笑, 代少升, 黄炼. 基于可解释深度学习的单通道脑电跨被试疲劳驾驶检测[J]. 仪器仪表学报, 2023, 44(5):140-149.
	FENG Xiao, DAI Shaosheng, HUANG Lian. Inter-Subject Fatigue Driving Detection Based on Interpretable Deep Learning with Single-channel EEG[J]. Journal of Instruments and Instrumentation, 2023, 44(5):140-149.
[22]	龚子安, 顾正晖, 陈迪. 基于局部与全局特征集成网络的跨被试驾驶疲劳检测(2024)[J/OL]. 计算机科学,[2024-10-23]. https://link.cnki.net/urlid/50.1075.tp.20240823.1505.006.
	GONG Zi’an, GU Zhenghui, CHEN Di. Cross-Subject Driving Fatigue Detection Based on Local and Global Feature Ensemble Network(2024)[J/OL]. Computer Science,[2024-10-23]. https://link.cnki.net/urlid/50.1075.tp.20240823.1505.006.
[23]	WANG L, JOHNSON D, LIN Y. Using EEG to Detect Driving Fatigue Based on Common Spatial Pattern and Support Vector Machine[J]. Turkish Journal of Electrical Engineering and Computer Sciences, 2021, 29(3):1429-1444.
[24]	杨利英, 孟天昊, 张清杨, 等. 特征融合实现脑电信号情感分析[J]. 西安电子科技大学学报, 2022, 49(6):95-102.
	YANG Liying, MENG Tianhao, ZHANG Qingyang, et al. Feature Fusion for EEG Signal Emotion Analysis[J]. Journal of Xidian University, 2022, 49(6):95-102.
[25]	LI Y, HE J. A Review of Strategies to Detect Fatigue and Sleep Problems in Aviation:Insights from Artificial Intelligence[J]. Archives of Computational Methods in Engineering, 2024, 31(8):4655-4672.
[26]	GRAMFORT A, LUESSI M, LARSON E, et al. MNE Software for Processing MEG and EEG Data[J]. Neuroimage, 2014, 86:446-460. doi: 10.1016/j.neuroimage.2013.10.027 pmid: 24161808
[27]	WANG Y, HAN M, PENG Y, et al. LGNet:Learning Local-Global EEG Representations for Cognitive Workload Classification in Simulated Flights[J]. Biomedical Signal Processing and Control, 2024,92:106046.
[28]	TAHERI GORJI H, WILSON N, VANBREE J, et al. Using Machine Learning Methods and EEG to Discriminate Aircraft Pilot Cognitive Workload During Flight[J]. Scientific Reports, 2023, 13(1):2507. doi: 10.1038/s41598-023-29647-0 pmid: 36782004
[29]	SEGU M, TONIONI A, TOMBARI F. Batch Normalization Embeddings for Deep Domain Generalization[J]. Pattern Recognition, 2023,135:109115.
[30]	翟凤文, 孙芳林, 金静. 多尺度卷积结合Transformer的抑郁脑电分类研究[J]. 西安电子科技大学学报, 2024, 51(2):182-195.
	ZHAI Fengwen, SUN Fanglin, JIN Jing. Research on Depression EEG Classification Using Multi-Scale Convolution Combined with Transformer[J]. Journal of Xidian University, 2024, 51(2):182-195.
[31]	TORSVALL L. Sleepiness on the Job:Continuously Measured EEG Changes in Train Drivers[J]. Electroencephalography and Clinical Neurophysiology, 1987, 66(6):502-511.
[32]	范晓丽, 牛海燕, 周前祥, 等. 基于EEG的脑力疲劳特征研究[J]. 北京航空航天大学学报, 2016, 42(7):1406-1413.
	FAN Xiaoli, NIU Haiyan, ZHOU Qianxiang, et al. Study on Mental Fatigue Features Based on EEG[J]. Journal of Beijing University of Aeronautics and Astronautics, 2016, 42(7):1406-1413.
[33]	BLANCO-DÍAZ C F, GUERRERO-MENDEZ C D, DELISLE-RODRIGUEZ D, et al. Evaluation of Temporal,Spatial and Spectral Filtering in CSP-Based Methods for Decoding Pedaling-Based Motor Tasks Using EEG Signals[J]. Biomedical Physics & Engineering Express, 2024, 10(3):035003.
[34]	DING Y, ROBINSON N, ZHANG S, et al. Tsception:Capturing Temporal Dynamics and Spatial Asymmetry from EEG for Emotion Recognition[J]. IEEE Transactions on Affective Computing, 2022, 14(3):2238-2250.
[35]	DING Y, ROBINSON N, TONG C, et al. LGGNet:Learning from Local-Global-Graph Representations for Brain-Computer Interface[J]. IEEE Transactions on Neural Networks and Learning Systems, 2023:1-14.
[36]	DING Y, LI Y, SUN H, et al. EEG-Deformer:A Dense Convolutional Transformer for Brain-Computer Interfaces[J]. IEEE Journal of Biomedical and Health Informatics, 2025, 29(3):1-14.
[37]	LAWHERN V J, SOLON A J, WAYTOWICH N R, et al. EEGNet:A Compact Convolutional Neural Network for EEG-Based Brain-Computer Interfaces[J]. Journal of Neural Engineering, 2018, 15(5):056013.

层类型	参数	输出尺寸	输出表示	参数数量
卷积层	8×1×64	8×31×256	FCT	512
批归一化		8×31×256		16
SEConv		8×31×256		140
卷积层	8×31×1	8×1×256		248
批归一化		8×1×256	F1T	16
激活函数	ELU	8×1×256		0
平均池化层	1×4	8×1×64	F1T/4	0
正则化层		8×1×64		156
SE		16×1×64	DF1T/4	64
卷积层	16×1×16	16×1×64		256
卷积层	16×1×1	16×1×64		256
批归一化		16×1×64		32
激活函数	ELU	16×1×64		0
平均池化层	1×8	16×1×8	DF1T/32	0
正则化层		16×1×8		352
全连接层		128		512
全连接层		2

软件/硬件	版本/配置
操纵系统	Windows10
处理器	Intel(R)Core(TM)i7-10700HX CPU@2.90GHz
内存	64 GB
编程语言	Python3.8.8
深度学习框架	Pytorch1.9.1

模型	实验训练
模型	准确率	召回率	F1值	特异性
SVM	53.05	51.21	56.43	80.87
KNN	53.88	69.33	59.94	85.40
RandomForest	61.77	59.68	59.13	78.49
SSENet	92.66	90.21	90.10	89.54

模型	实验训练
模型	准确率	召回率	F1值	特异性
CSP+SVM	86.03	70.02	86.41	86.09
CSP+KNN	80.23	76.56	80.67	80.03
CSP+ RandomForest	82.61	73.46	82.67	82.42
SSENet	92.66	90.21	90.10	89.54

模型	实验训练
模型	模型来源	第一折	第二折	第三折	第四折	第五折	均值
Tsception	文献[34]	89.02	91.66	86.11	90.27	87.63	88.94
LGGNet	文献[35]	69.86	74.16	75.83	77.77	72.22	73.97
EEG-Deformer	文献[36]	82.77	83.47	79.58	84.16	86.25	83.25
EEGNet	文献[37]	89.64	86.60	86.69	88.75	88.92	88.17
SSENet	文中模型	92.50	92.91	92.22	93.19	92.50	92.66