Crosswind landing is one of the important training subjects for flight trainees in the intermediate stage of flight training. Accurately assessing the workload of trainees during this training is essential for mitigating human operational risks and enhancing the overall quality and effectiveness of the training process.To address this requirement, a non-invasive method for predicting the workload of flight trainees was proposed, which integrated facial action units (AUs), gaze data, and head posture data. First, a flight training scenario for crosswind landings was constructed. Based on this scenario, the factors that influence workload within the “human-machine-environment” system were identified.Subsequently, a comprehensive evaluation system for the workload during crosswind landings was established, integrating both subjective and objective data. Building on this foundation, a flight experiment using a crosswind landing simulator was designed to collect objective data on trainees' AUs, gaze, and head posture, along with subjective workload assessments based on the NASA-TLX scale.The CRITIC model was then employed to determine the feature weights of the six dimensions in the NASA-TLX scale. According to the levels of workload, flight trainees were categorized into three groups: low, medium, and high. The symmetric dot pattern (SDP) was utilized to perform a nonlinear mapping of the objective data into the polar coordinate space, enabling effective data fusion.Furthermore, fuzzy C-means clustering analysis was applied to explore the correlation between the fused data and subjective workload, leading to the generation of a fused feature pattern that measures workload and its dimensions.Finally, based on this feature pattern, four machine learning algorithms, including support vector machine (SVM), random forest, logistic regression, and extreme gradient boosting, were employed to construct a workload prediction model for flight trainees during crosswind landing. The experimental results demonstrate that the workload prediction model based on the “SVM + fusion feature pattern” achieves the highest accuracy, effectively measuring the workload of trainees in crosswind landing flight training.
HE Qiang, PENG Han-lu, GU Yu-tong.
Crosswind Landing Workload Prediction for Flight Trainees Based on Machine Learning[J].
Science Technology and Engineering, 2026, 26(11): 4864-4876 DOI:10.12404/j.issn.1671-1815.2503516
综上所述,结合主观评价和多模态生理数据的飞行工作负荷预测成为飞行人为因素研究的主流[24-25],但多模态生理数据依赖可穿戴设备或贴片组件,造成了任务入侵,可能影响飞行学员的本体感受。特别是,现有研究多集中于飞行任务的总体工作负荷评价,缺乏对构成工作负荷的维度的预测评价,以致不能追溯影响飞行学员工作负荷管理的“人-机-环境”系统因素,从而不能生成更优的工作负荷管理和有效的着陆训练策略。为此,现以中教阶段的大侧风着陆训练为研究对象,利用大侧风着陆训练中飞行学员的面部视频,构建基于面部活动单元(action unit, AU)、注视类与头部姿态类数据融合的着陆飞行工作负荷评价方法。通过分析大侧风着陆飞行训练场景,明确着陆飞行工作负荷综合评测体系,进而设计实施模拟机大侧风着陆实验,采集相关数据,基于数据分析提取度量工作负荷及其6个评价维度的客观特征模式;最后,基于该特征模式,结合机器学习算法构建工作负荷预测模型。
FengC Y, WanyanX R, YangK, et al. A comprehensive prediction and evaluation method of pilot workload[J]. Technology and Health Care, 2018, 26: 65-78.
[2]
McfarlanR A. Human factors in air transport design[M]. New York: McGraw-Hill, 1946.
[3]
KellyD, EfthymiouM. An analysis of human factors in fifty controlled flight into terrain aviation accidents from 2007 to 2017[J]. Journal of Safety Research, 2019, 69(6): 155-165.
[4]
BoeingCompany. Virtual reality training for crosswind landing: a study on cognitive load and performance metrics: BCA-TR-2021-015[D]. Seattle: Boeing Commercial Airplanes, 2021.
[5]
ParasuramanR, SheridanT B, WickensC D. A model for types and levels of human interaction with automation[J]. Human Factors, 2000, 42(3): 367-381.
[6]
董大勇. 民机适航审定中的机组工作量测量方法研究[D]. 上海: 上海交通大学, 2012.
[7]
DongDayong. Research on measurement methods of crew workload in civil aircraft airworthiness certification[D]. Shanghai: Shanghai Jiao Tong University, 2012.
[8]
WickensC D. Multiple resource theory and workload assessment[J]. Human Factors, 2002, 44(3): 449-455.
[9]
WeiZ M, ZhuangD M, WanyanX R, et al. A model for discrimination and prediction of mental workload of aircraft cockpit display interface[J]. Chinese Journal of Aeronautics, 2014, 27(5): 1070-1077.
[10]
LandryS J, LaguA, KinnariJ. State-based modeling of continuous human-integrated systems: an application to air traffic separation assurance[J]. Reliability Engineering & System Safety, 2010, 95(4): 345-353.
[11]
HartS G, StavelandL E. Development of NASA-TLX (task load index): results of empirical and theoretical research[J]. Advances in Psychology, 1988, 52: 139-183.
[12]
LinC J, LinP H, ChenH J, et al. Effects of controller-pilot communication medium, flight phase and the role in the cockpit on pilots’ workload and situation awareness[J]. Safety Science, 2012, 50(9): 1722-1731.
[13]
YanS, CongC T, ChenY, et al. Effect of user interface layout on the operators’ mental workload in emergency operating procedures in nuclear power plants[J]. Nuclear Engineering & Design, 2017, 322: 266-276.
[14]
WangY, ZhangC, LiuC, et al. Analysis on pulse rate variability for pilot workload assessment based on wearable sensor[J]. Human Factors & Ergonomics in Manufacturing & Service Industries, 2024, 34(6): 1-12.
[15]
JiangY, ChengB. Workload assessment in aviation using NASA-TLX[J]. Ergonomics, 2023, 65(4): 547-560.
XinTian, LuoZhiyan, JiangLitian, et al. Analysis of pilots' workload during aircraft landing phase[J]. Human Factors and Ergonomics in Manufacturing & Service Industries, 2019, 25(2): 62-65.
[18]
WangX S, GongG H, LiN, et al. Use of multimodal physiological signals to explore pilots' cognitive behavior during flight strike task performance[J]. Medicine in Novel Technology and Devices, 2020, 5: 1-6.
WeiZongmin, HaoHongxun, XuQizhi, et al. Advances in measurement indicators and evaluation methods of pilot mental workload[J]. Science Technology and Engineering, 2019, 19(24): 1-8.
[21]
MohanaveluK, PoonguzhaliS, JananiA, et al. Machine learning-based approach for identifying mental workload of pilots[J]. Biomedical Signal Processing and Control, 2022, 75: 103623.
[22]
ChenJ, XueL, RongJ, et al. Real-time evaluation method of flight mission load based on sensitivity analysis of physiological factors[J]. Chinese Journal of Aeronautics, 2022, 35(3): 450-463.
WangWei, ZhaoMinrui, GaoHongni, et al. Intent recognition in human-computer interaction based on EEG and eye-movement signals[J]. Chinese Journal of Aeronautics, 2021, 42(2): 324290.
[25]
DebieE, RojasR F, FidockJ, et al. Multimodal fusion for objective assessment of cognitive workload: a review[J]. IEEE Transactions on Cybernetics, 2021, 51(3): 1542-1555.
LiuZhongqi, YuanXiugan, FanYubo. Flight performance evaluation model based on BP neural network[J]. Journal of Beijing University of Aeronautics and Astronautics, 2010, 36(4): 403-406.
MaFei, ZhangQiong, LaiPeijun, et al. Quantitative safety model for test flight training based on BP neural network[J]. Chinese Journal of Aeronautics, 2024, 45(5): 409-422.
[30]
TaoD, TanH, WangH, et al. A systematic review of physiological measures of mental workload[J]. International Journal of Environmental Research and Public Health, 2019, 16(15): 2716.
ZhengLing, QiaoXuqiang, NiTao, et al. Study on driver cognitive load based on multidimensional information feature analysis[J]. China Journal of Highway and Transport, 2021, 34(4): 11-21.
[33]
NiuW K, JiangY Y, ZhangX, et al. Changes of effective connectivity in the alpha band characterize differential processing of audiovisual information in cross-modal selective attention[J]. Neuroscience Bulletin, 2020, 36(9): 1009-1022.
FengChuanyan, WanyanXiaoru, LiuShuang, et al. Study on situation awareness under different mental workloads[J]. Journal of Northwestern Polytechnical University, 2020, 38(3): 610-618.
[36]
DiakoulakiD, MavrotasG, PapayannakisL. Determining objective weights in multiple criteria problems: the critic method[J]. Computers & Operations Research, 1995, 22(7): 763-770.
[37]
YangL, DengM. Based on K-means and fuzzy K-means algorithm classification of precipitation[C]// International Symposium on Computational Intelligence and Design. Hangzhou: IEEE, 2010: 218-221.
[38]
JiddahS M, YurtkanK. Feature fusion for human compound emotion recognition: a fusion of facial expression texture and action unit data[J]. Pattern Analysis and Applications, 2024, 27(4): 1-15.
PanPengcheng, XiangYang, ChenTianyou, et al. A motor fault diagnosis method based on multi-channel data fusion and CNN[J]. Journal of Ship and Ocean Engineering, 2024, 53(4): 29-35.
[42]
ImantsP, GreefT D. Using eye tracker data in air traffic control[C]// European Conference on Cognitive Ergonomics. New York: ACM, 2011: 259-260.
[43]
ChauhanV K, DahiyaK, SharmaA, et al. Problem formulations and solvers in linear SVM: a review[J]. Artificial Intelligence Review, 2018, 52(2): 803-855.
[44]
JalalN, MehmoodA, ChoiG S, et al. A novel improved random forest for text classification using feature ranking and optimal number of trees[J]. Journal of King Saud University-Computer and Information Sciences, 2022, 34(6): 2733-2742.
[45]
ChristodoulouE, MaJ, CollinsG S, et al. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models[J]. Journal of Clinical Epidemiology, 2019, 110: 12-22.
[46]
ChenT, GuestrinC. XGBoost:a scalable tree boosting system[C]// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD ’16). New York: ACM, 2016: 785-794.
[47]
Federal Aviation Administration. Crew resource management training: AC 120-51E[R]. Washington, DC: Federal Aviation Administration, 2004.