众所周知，当人长期处于压力状态下时，会对人的心理、生理、甚至行为等多方面产生不利的影响。煤矿生产属于危险系数较高的行业，矿工在工作中的压力状态更应该值得关注，对此进行识别，并有针对性的进行干预，可有效管控矿工的不安全行为，降低煤矿安全管理难度。本研究通过设计脑电实验、采集并统计分析了被试在低、中和高三种工作压力任务下的脑电信号、行为学数据和主观压力得分，提取出能够反映工作压力变化的指标，在此基础上构建基于机器学习的矿工工作压力识别模型，对矿工的工作压力等级进行评估，实现及时准确地预判矿工的工作状态，进而完善煤矿安全管理，主要研究内容和结论如下：

（1）设计并完成了工作压力诱发的脑电实验。采集了被试分别在低、中和高工作压力任务下的脑电指标和行为学指标（平均反应时间、正确率），使用美国航天局任务负荷指数（NASA-TLX）量表获取被试的主观工作压力得分。使用SPSS26.0对行为学指标和NASA-TLX得分进行统计分析，基于NASA-TLX得分验证了本文工作压力诱发实验的有效性。通过统计分析行为学指标，发现当工作压力任务难度增加时，被试的反应时间和失误量均显著增加。

（2）分析脑电数据，确定了可以表征工作压力的脑电指标。使用快速傅里叶变换法提取θ、α和β波段在低、中和高工作压力任务下的功率和功率谱密度值，根据地形图和配对样本t检验结果，随着工作压力任务难度的上升，θ和β波段的功率谱密度值在大脑多个区域出现了显著性增加，α波段的功率谱密度值在大脑多个区域出现了显著性降低。根据低、中和高工作压力任务的脑电指标：α/（α+β+θ）、β/（α+β+θ）和α/β的功率值显著性检验结果，不同工作压力任务下，每位被试均在FP1、F7、F3、O1、OZ五个通道具有显著差异性。对低、中和高工作压力任务下的脑电指标和NASA-TLX得分进行相关性分析，发现脑电指标α/（α+β+θ）和α/β在F7、O1和OZ通道与NASA-TLX得分显著负相关，而脑电指标β/（α+β+θ）在F7、O1和OZ通道与NASA-TLX得分显著正相关。

（3）使用K-means聚类分析法对工作压力程度进行了分级。对实验所获得的行为学数据、NASA-TLX得分和脑电指标进行聚类，根据聚类结果，可将矿工的工作压力等级划分为三类，分别为：正常、中度工作压力和重度工作压力。

（4）基于机器学习建立了矿工工作压力识别模型。将筛选出来的矿工工作压力表征指标作为输入向量，工作压力等级作为输出向量，基于支持向量机、K近邻和BP神经网络算法分别构建矿工工作压力识别模型。通过计算并综合对比三种识别算法的总体识别正确率、精确度、召回率、F1值，发现K近邻算法整体最优，其识别正确率、精确度、召回率、F1值分别为：95.83%、95.77%、96.30%、95.84%。故本研究选择K近邻算法对矿工的工作压力进行识别，为后续煤矿安全管理和搭建矿工工作压力预警平台提供理论依据。

It is well known that when people are under stress for a long time, it will have adverse effects on their psychological, physiological, and even behavioral aspects. As coal mine production is an industry with a high-risk factor, the stressful state of miners at work should be more worthy of attention. Identifying this and making targeted interventions can effectively control miners' unsafe behaviors and reduce the difficulty of coal mine safety management. In this study, we designed EEG experiments, collected and statistically analyzed EEG signals, behavioral data, and subjective stress score of subjects under three types of work stress tasks: low, medium, and high, extracted indicators that reflect changes in work stress, and built a machine learning-based work stress recognition model for miners to evaluate their work stress levels and achieve timely and accurate prediction of miners' The main research contents and conclusions are as follows:

(1) An EEG experiment induced by work stress was designed and completed. EEG and behavioral indicators (average reaction time, correctness) were collected from subjects under low, medium, and high work stress tasks, respectively, and subjective work stress scores were obtained from subjects using the National Aeronautics and Space Administration-Task Load Index(NASA-TLX)task scale. Statistical analysis of behavioral indicators and NASA-TLX scores was performed using SPSS 26.0, and the validity of the work stress-induced experiment in this paper was verified based on NASA-TLX scores. By statistically analyzing the behavioral indicators, it was found that when the difficulty of the job stress task increased, the subjects' reaction time and the number of errors increased significantly.

(2) EEG data were analyzed to identify EEG indicators that could characterize work stress. The power and power spectral density values of theta, alpha, and beta bands were extracted using the fast Fourier transform method for low, medium, and high work-stress tasks. According to the topography and paired sample t-test results, as the difficulty of the work stress task increased, the power spectral density values of theta and beta bands showed a significant increase in several regions of the brain, and the power spectral density values of the alpha band showed a significant decrease in several regions of the brain. According to the results of the significance tests of the power values of the EEG indicators: α/(α+β+θ), β/(α+β+θ), and α/β for the low, medium, and high work stress tasks, each subject had significant differences in the five channels of FP1, F7, F3, O1, and OZ under different work stress tasks. Correlation analysis of EEG indicators and NASA-TLX scores under low, medium, and high work stress tasks revealed that EEG indicators α/(α+β+θ) and α/β were significantly negatively correlated with NASA-TLX scores in F7, O1, and OZ channels, while EEG indicators β/(α+β+θ) were significantly positively correlated with NASA-TLX scores in F7, O1 and OZ channels.

(3) The degree of work stress was graded using K-means cluster analysis. The behavioral data, NASA-TLX scores, and EEG indicators obtained from the experiment were clustered, and based on the clustering results, the work stress levels of miners could be classified into three categories: normal, moderate work stress and severe work stress, respectively.

(4) A miners' work stress identification model was established based on machine learning. The filtered work pressure characterization indexes of miners were used as input vectors and work pressure levels were used as output vectors to build miners' work pressure recognition models based on support vector machine, K-nearest neighbor, and BP neural network algorithms, respectively. By calculating and comparing the overall recognition accuracy, precision, recall, and F1 values of the three algorithms, it is found that the K nearest neighbor algorithm is the best overall, and its recognition accuracy, precision, recall, and F1 values are 95.83%, 95.77%, 96.30%, and 95.84%, respectively. Therefore, this study selects the K-nearest neighbor algorithm to identify the work stress of miners, which provides a theoretical basis for the subsequent coal mine safety management and the construction of a work stress warning platform for miners.

[1] 袁亮. 我国煤炭主体能源安全高质量发展的理论技术思考[J]. 中国科学院院刊, 2023, 38(01): 11-22.

[2] 秦波涛, 蒋文婕, 史全林, 等. 矿井粉煤灰基防灭火技术研究进展[J]. 煤炭科学技术, 2023, 51(01): 329-342.

[3] 刘银奎. 基于机器视觉的井下人脸识别系统设计[D]. 徐州: 中国矿业大学, 2022.

[4] 许鹏飞. 2000—2021年我国煤矿事故特征及发生规律研究[J]. 煤炭工程, 2022, 54(07): 129-133.

[5] 田水承, 景国勋. 安全管理学[M]. 北京: 机械工业出版社, 2022.

[6] 陈红, 祁慧, 谭慧. 基于特征源与环境特征的中国煤矿重大事故研究[J]. 中国安全科学学报, 2005, (09): 33-38+115.

[7] Miller R L, Griffin M A, Hart P M. Personality and Organizational Health: The Role of Conscientiousness [J]. Work & Stress, 1999, 13(1): 7-19.

[8] 田水承, 匡秘姈, 丁洋. 风险偏好中介作用下矿工工作压力对不安全行为的影响[J]. 安全与环境学报, 2023, 23(01): 132-138.

[9] 田水承, 蔡欣甫, 赵钊颖, 等. 矿工不安全情绪状态管理演化博弈研究[J/OL]. 矿业安全与环保, 1-6.

[10] 王璟. 矿工安全心理资本与不安全行为的关系研究[D]. 西安: 西安科技大学, 2016.

[11] 叶新凤, 李新春, 王智宁. 安全氛围, 工作压力与安全行为[J]. 技术经济与管理研究, 2014, (10): 44-50.

[12] Cannon W. The Problem of Miracles in the 1830 ' s[J]. Victorian Studies, 1960, 4(1): 5-32.

[13] Selye H. Stress Situations[J]. American Journal of Psychiatry, 1955, 112(4).

[14] French J R P, Kahn R L. A Programmatic Approach to Studying the Industrial Environment and Mental Health[J]. Industrial Organizations and Health, 2013, 18(3): 1-47.

[15] Jex S M, Gudanowski D M. Efficacy Beliefs and Work Stress: An Exploratory Study[J]. Journal of Organizational Behavior, 1992, 13(5): 509-517.

[16] Akerstedt T, Fredlund P, Gillberg M, et al. Work Load and Work Hours in Relation to Disturbed Sleep and Fatigue in a Large Representative Sample[J]. Journal of Psychosomatic Research, 2002, 53(1): 585-588.

[17] Semmer N K, Meier L L. Individual Differences, Work Stress and Health[J]. 2009, 83-120.

[18] Himle D P, Jayaratne S D, Chess W A. Gender Differences in Work Stress Among Clinical Social Workers[M]. Burnout among Social Workers. Routledge. 2013, 10(1): 41-56.

[19] 徐金芹. 领导干部心理压力产生的性格原因及应对策略[J]. 领导科学, 2013(34): 37-39.

[20] 田水承, 孔维静, 况云, 等. 矿工心理因素、工作压力反应和不安全行为关系研究[J]. 中国安全生产科学技术, 2018, 14(08): 106-111.

[21] 郭楠, 王茜, 石坤杰, 等. 煤矿安全管理人员工作压力相关研究[J]. 中国安全科学学报, 2020, 30(5): 1-6.

[22] 沈玉志, 丁天维. 工作压力对制造企业一线员工安全绩效的影响机制[J]. 安全与环境学报, 2021, 21(06): 2638-2645.

[23] 侯瑞, 薛文星, 何小红, 等. 社区医疗机构护理人员的心理健康状况及与工作压力的关系[J]. 健康研究, 2022, 42(5): 517-520.

[24] 陈岩. 足球裁判员初学者心理压力的形成原因与对策[J]. 运动精品, 2022, 41(2): 107-109.

[25] Matteson M T, Ivancevich J M. Controlling Work Stress: Effective Human Resource and Management Strategies[M]. Jossey-Bass, 1987, 14(3): 460-482.

[26] Peterson M, Wilson J F. The Culture-Work-Health Model and Work Stress[J]. American Journal of Health Behavior, 2002, 26(1): 16-24.

[27] Wainwright D, Calnan M. Work stress: The making of a Modern Epidemic[M]. McGraw-Hill Education (UK), 2002, 31(6): 1282-1283.

[28] 植凤英, 张进辅. 西南少数民族民众心理压力特点及其与社会比较的关系[J]. 西南大学学报(自然科学版), 2012, 34(10): 150-156.

[29] 张娇. 基于SEM的矿工工作压力对煤矿险兆事件的影响研究[J]. 煤矿安全, 2018, 49(08): 285-288.

[30] 徐文翠, 刘明曦, 任霄凌, 等. 三甲医院ICU护士工作压力与社会支持的相关性研究[J]. 中国医院管理, 2018, 38(08): 70-71+80.

[31] Noeeis L A, Didymus F F, Kaiseler M. How can Social Support Make Coaching Less Stressful? A Longitudinal Inquiry With Sports Coaches[J]. International Journal of Sports Science & Coaching, 2022, 17(6): 1258-1269.

[32] 郭进平, 孙寅峰, 马卓远, 等. 代际下工作压力对矿工行为影响: 差错管理氛围的中介作用[J]. 中国矿业, 2022, 31(12): 21-28.

[33] 李正东, 李家成. 工作家庭的边界弹性对工作压力的影响机制研究[J]. 中国人事科学, 2022(08): 64-80.

[34] 李正东, 李家成. 职业兴趣获得对工作压力的缓解效应与干预机制[J]. 当代青年研究, 2022(04): 43-54+62.

[35] 刘静芳, 宫璇, 蔡忠香. 湖北省5家三级甲等综合医院精神科护士知觉压力、职业倦怠与心理健康的现状及相关性[J]. 职业与健康, 2021, 37(07): 927-930+935.

[36] 于路. 基于心电指标的心理压力检测研究[J]. 心理科学, 2017, 40(02): 277-282.

[37] 李烨, 杨英, 汪偌宁, 等. 新冠病毒感染疫情期间北京市临床住院医师工作压力与职业倦怠状况调查[J]. 医学与社会, 2023, 36(04): 122-126.

[38] 李昕, 张云鹏, 李红红, 等. 针对个体差异的心理压力评估[J]. 中国生物医学工程学报, 2014, 33(01): 45-50.

[39] 左娟. 期望对工作压力调节关系的实验研究[D]. 成都: 西南交通大学, 2017.

[40] 赵湛, 韩璐, 方震, 等. 基于可穿戴设备的日常压力状态评估研究[J]. 电子与信息学报, 2017, 39(11): 2669-2676.

[41] 杨妮. 基于HRV的心理压力识别算法研究与实现[D]. 西安: 西安工业大学, 2019.

[42] Hosseini S A, Khalizadeh M A. Emotional Stress Recognition System Using EEG and Psychophysiological Signals: Using New Labelling Process of EEG Signals in Emotional Stress State[C]// 2010 international conference on biomedical engineering and computer science. IEEE, 2010: 1-6.

[43] Singh R R, Conjeti S, Banerjee R. A Comparative Evaluation of Neural Network Classifiers for Stress Level Analysis of Automotive Drivers Using Physiological Signals[J]. Biomedical Signal Processing and Control, 2013, 8(6): 740-754.

[44] Hsieh C-P, Chen Y-T, BEH W-K, et al. Feature Selection Framework for XGBoost Based on Electrodermal Activity in Stress Detection[C]// 2019 IEEE International Workshop on Signal Processing Systems (SiPS). IEEE, 2019: 330-335.

[45] Su C. Heart Rate Variability Feature Selection Using Random Forest for Mental Stress Quantification[D]. Montreal: Concordia University, 2020.

[46] Fan X-A, Bi L-Z, Chen Z-L. Using EEG to Detect Drivers ' Emotion with Bayesian Networks[C]// 2010 International Conference on Machine Learning and Cybernetics. IEEE, 2010, 3: 1177-1181.

[47] 叶朋飞, 陈兰岚, 张傲. 基于禁忌搜索的混合算法在驾驶压力识别中的应用[J]. 华东理工大学学报(自然科学版), 2018, 44(05): 730-736.

[48] 张思美, 王海鹏, 刘栋, 等. 智能驾驶中的行为辅助压力感知方法[J]. 软件学报, 2019, 29(S2): 86-95.

[49] 尚宇, 杨妮. 改进粒子群优化 BP 神经网络的心理压力识别算法[J]. 科学技术与工程, 2020, 20(04): 1467-1472.

[50] 江润强, 陈兰岚, 谌鈫. 基于单模态生理信号无监督特征学习的驾驶压力识别[J]. 华东理工大学学报(自然科学版), 2021, 47(04): 475-482.

[51] 林艳飞, 龙媛, 张航, 等. 基于XGBoost的多种生理信号评估心理压力等级方法[J]. 北京理工大学学报自然版, 2022, 42(8): 871-880.

[52] Hayashi H, Tsuji T. Human–Machine Interfaces Based on Bioelectric Signals: A Narrative Review with a Novel System Proposal[J]. IEEJ Transactions on Electrical and Electronic Engineering, 2022, 17(11): 1536-1544.

[53] 田水承, 胥静, 田方圆, 等. 基于EEG的夜班矿工疲劳检测[J]. 安全与环境学报, 2022, 22(04): 2034-2040.

[54] Perris C, Monakhov K. Depressive Symptomatology and Systemic Structural Analysis of the EEG[J]. Hemisphere Asymmetries of Function in Psychopathology, 1979, 223-236.

[55] Perris C. Averaged Evoked Responses (are) in Patients with Affctive Disoredrs: A Pilot Study of Possible Hemispheric Differences in Depressed Patients[J]. Acta Psychiatrica Scandinavica, 1974, 50: 89-98.

[56] Matsunami K, Homma S, Han X Y, et al. Generator Sources of EEG Large Waves Elicited by Mental Stress of Memory Recall or Mental Calculation[J]. The Japanese Journal of Physiology, 2001, 51(5): 621-624.

[57] Lewis R S, Weekes N Y, Wang T H. The Effect of a Naturalistic Stressor on Frontal EEG Asymmetry, Stress, and Health[J]. Biological Psychology, 2007, 75(3): 239-247.

[58] Li H-C O, Seo J, Kham K, et al. Measurement of 3D Visual Fatigue Using Event-Related Potential (ERP): 3D Oddball Paradigm[C]// 2008 3DTV conference: the true vision-capture, transmission and display of 3D video. IEEE, 2008: 213-216.

[59] Zhao W, Liu L, Zheng F, et al. Investigation into Stress of Mothers with Mental Retardation Children Based on EEG (Electroencephalography) and Psychology Instruments[C]// Brain Informatics: International Conference, BI 2011, Lanzhou, China, September 7-9, 2011. Proceedings. Springer Berlin Heidelberg, 2011: 238-249.

[60] Peng H, Hu B, Zheng F, et al. A Method of Identifying Chronic Stress by EEG[J]. Personal and Ubiquitous Computing, 2013, 17(7): 1341-1347.

[61] 李昕, 李红红, 李长吾. 基于复杂度熵特征融合的高压力人群情感状态评估[J]. 中国生物医学工程学报, 2013, 32(03): 313-320.

[62] 柳长源, 李文强, 毕晓君. 基于脑电信号的情绪特征提取与分类[J]. 传感技术学报, 2019, 32(01): 82-88.

[63] 王大颜, 瞿珏. 基于脑电指标的指控系统人机界面评价研究[J]. 兵器装备工程学报, 2021, 42(11): 196-203.

[64] 祁首铭. 不同认知负荷下考虑驾驶人生理特征的行车风险辨识方法[D]. 哈尔滨: 哈尔滨工业大学, 2021.

[65] Lazarus R S, Launier R. Stress-Related Transactions between Person and Environment[J]. Perspectives in Interactional Psychology, 1978, 287-327.

[66] Summers T P, Decotiis T A, Denisi A S. A Field Study of Some Antecedents and Consequences of Felt Job Stress[M]. Occupational Stress. CRC Press. 2020, 113-128.

[67] Mcgrath J E. Stress and Behavior in Organizations[J]. Handbook of Industrial and Organizational Psychology, 1976, 1351-1396.

[68] Morrison E W, Milliken F J. Organizational silence: A Barrier to Change and Development in a Pluralistic World[J]. Academy of Management Review, 2000, 25(4): 706-725.

[69] 田水承, 郭彬彬, 李树砖. 煤矿井下作业人员的工作压力个体因素与不安全行为的关系[J]. 煤矿安全, 2011, 42(09): 189-192.

[70] Kulic L, Galjak M, Jovanovic J. Correlation of Socio Demographic Variables and Interpersonal Sources of Stress at Work among Miners[J]. Vojnosanitetski Pregled, 2019, 76(11): 1169-1177.

[71] 李琰, 张燕. PSB与工作压力、工作资源及PSC的关系研究[J]. 中国安全科学学报, 2019, 29(01): 1-6.

[72] 李乃文, 刘孟潇, 牛莉霞. 辱虐管理对安全绩效的影响——心理痛苦和心智游移的链式中介作用[J]. 软科学, 2019, 33(09): 60-63.

[73] 庞晓华, 栗继祖. 基于有调节中介模型的组织公平感与矿工不安全行为的关系[J]. 煤矿安全, 2021, 52(10): 256-260.

[74] Le A B, Balogun A O, Smith T D. Long Work Hours, Overtime, and Worker Health Impairment: A Cross-Sectional Study among Stone, Sand, and Gravel Mine Workers[J]. International Journal of Environmental Research and Public Health, 2022, 19(13).

[75] Daud S N S S, Sudirman R. Wavelet Based Filters for Artifact Elimination in Electroencephalography Signal: A Review[J]. Annals of Biomedical Engineering, 2022, 50(10): 1271-1291.

[76] Bos D O. EEG-Based Emotion Recognition[J]. The Influence of Visual and Auditory Stimuli, 2006, 56(3): 1-17.

[77] 赵文. 基于脑电信号的特定人群心理压力研究[D]. 兰州: 兰州大学, 2012.

[78] Arsalan A, Majid M, Butt A R, et al. Classification of Perceived Mental Stress Using a Commercially Available EEG Headband[J]. IEEE Journal of Biomedical and Health Informatics, 2019, 23(6): 2257-2264.

[79] 邢毅超. 虚拟与现实单色光环境下人体生理心理应激反应研究[D]. 北京: 北京建筑大学, 2022.

[80] 陈小刚, 赵秉, 刘明, 等. 稳态视觉诱发电位脑-机接口控制机械臂系统的设计与实现[J]. 生物医学工程与临床, 2018, 22(03): 244-250.

[81] 龙飞, 吴小培, 范羚. 基于独立分量分析的脑电噪声消除[J]. 生物医学工程学杂志, 2003(03): 479-483.

[82] 赵仑. ERPs实验教程[J]. 东南大学出版社, 2010.

[83] Katmah R, Al-shargie F, Tariq U, et al. A Review on Mental Stress Assessment Methods Using EEG Signals[J]. Sensors, 2021, 21(15).

[84] Ameera A, Saidatul A, Ibrahin Z. Analysis of EEG Spectrum Bands Using Power Spectral Density for Pleasure and Displeasure State[C]// IOP Conference Series: Materials Science and Engineering, IOP Publishing, 2019, 557(1): 12-30.

[85] Umer Saeed S M, Anwer S M, Majid M, et al. Selection of Neural Oscillatory Features for Human Stress Classification with Single Channel EEG Headset[J]. BioMed research international, 2018: 1-8.

[86] 刘帅, 乌日开西·艾依提. 基于VMD与CSP的脑电特征提取方法[J]. 计算机仿真, 2022, 39(11): 432-437.

[87] 孟凡强. 基于数据挖掘的采煤工作面安全预警研究[D]. 北京: 北京科技大学, 2022.

[88] Karthikeyan P, Murugappan M, Yaacob S. A review on stress inducement stimuli for Assessing Human Stress Using Physiological Signals[C]// 2011 IEEE 7th International Colloquium on Signal Processing and its Applications. IEEE, 2011: 420-425.

[89] 张丽平, 詹长安. 心算任务复杂度对脑电 theta, alpha和beta波的影响[J]. 航天医学与医学工程, 2019, 32(3): 235-242.

[90] Schmidt P, Reiss A, Dvrichen R, et al. Wearable-based affect recognition—A review[J]. Sensors, 2019, 19(19).

[91] 于璐瑶, 田丽艳, 伍锡昌, 等. 基于独立成分分析的马里亚纳弧后地幔源区特征研究[J]. 海洋科学进展, 2021, 39(01): 73-88.

[92] Delorme A, Sejnowski T, Makeig S. Enhanced Detection of Artifacts in EEG Data Using Higher-Order Statistics and Independent Component Analysis[J]. Neuroimage, 2007, 34(4): 1443-1449.

[93] Al-shargie F, Tang T B, Badruddin N, et al. Simultaneous Measurement of EEG-fNIRS in Classifying and Localizing Brain Activation to Mental Stress [C]// 2015 IEEE International Conference on Signal and Image Processing Applications (ICSIPA). IEEE, 2015: 282-286.

[94] Al-shargie F, Tang T B, Badruddin N, et al. Mental Stress Quantification Using EEG Signals [C]// International Conference for Innovation in Biomedical Engineering and Life Sciences: ICIBEL2015, 6-8 December 2015, Putrajaya, Malaysia 1. Springer Singapore, 2016: 15-19.

[95] Akella A, Singh A K, Leong D, et al. Classifying Multi-Level Stress Responses from Brain Cortical EEG in Nurses and Non-health Professionals Using Machine Learning Auto encoder[J]. IEEE Journal of Translational Engineering in Health and Medicine, 2021, 9, 1-9.

[96] 郑霞忠, 张蒙, 陈云, 等. 基于眼-脑生理指标的大型地下洞室驾驶员疲劳演化特征研究[J/OL]. 安全与环境学报, 1-12.

[97] Lim H-K. Influence of Time Stress on EEG Characteristics Related with Human Errors[J]. Journal of the Korean Society of Safety, 2011, 26(3): 83-90.

[98] Awang S A, Pandiyan P M, Yaacob S, et al. Spectral Density Analysis: Theta Wave as Mental Stress Indicator [C]// Signal Processing, Image Processing and Pattern Recognition: International Conference, SIP 2011, Held as Part of the Future Generation Information Technology Conference FGIT 2011, in Conjunction with GDC 2011, Jeju Island, Korea, December 8-10, 2011. Proceedings. Springer Berlin Heidelberg, 2011: 103-112.

[99] Subhani A R, Xia L, Malik A S. Evaluation of Mental Stress Using Physiological Signals [C]// 2011 National Postgraduate Conference. IEEE, 2011: 1-4.

[100] Jensen O, Tesche C D. Frontal Theta Activity in Humans Increases with Memory Load in a Working Memory Task[J]. European Journal of Neuroscience, 2002, 15(8): 1395-1399.

[101] Harmony T, Fernandez T, Silva J, et al. EEG Delta Activity: an Indicator of Attention to Internal Processing During Performance of Mental Tasks[J]. International Journal of Psychophysiology, 1996, 24(1-2): 161-171.

[102] 万憬, 杨柳, 徐珀, 等. 急性高空缺氧对高性能战机飞行员基本认知能力的影响[J]. 军事医学, 2011, 35(12): 954-955.

[103] Rumelhart D E, Hinton G E, Williams R J. Learning Representations by Back-Propagating Errors[J]. Nature, 1986, 323(6088): 533-536.