2 State Key Laboratory of Air Traffic Management System, Beijing 100083, China
3 95861, Jiuquan 735300, China
Show less
文章历史+
Received
Published
Revised
2025-05-09
2026-07-08
2026-05-11
Online Date
2026-08-06
PDF (7691K)
摘要
为解决Transformer类算法在处理飞行航迹短期预测问题时,因固有Token计算策略导致多空间维度参数动态关联冗余计算、进而降低预测精度的实际问题,通过深入分析航迹数据的时空特性与Transformer类算法的计算逻辑,在对称多尺度骨干架构基础上,创新性地引入多维度拆分模块、局域特征提取模块与全局模式拟合模块,提出一种融合全局与局域特征的多维度计算飞行航迹短期预测算法。该算法将短期航迹预测任务映射为等长已知航迹序列到未知序列的模式回归问题,有效捕捉各维度数据的局部时空动态特征与全局时空依赖关系,进一步提升模型对复杂航迹模式的适应性与鲁棒性。基于OpenSky数据集中的海量真实航迹数据,分别开展对比实验与消融实验,结果表明:相较于最新Transformer类基准算法,该方法在平均绝对误差(mean absolute error,MAE)、均方根误差(root mean square error,RMSE)和平均绝对百分比误差(mean absolute percentage error,MAPE)指标上分别实现10.92%、14.39%和14.96%的性能提升;同时,逐模块消融验证显示算法准确性随模块集成逐步增强,充分证实所提算法的有效性、先进性及各模块对整体性能的贡献。
Abstract
In order to address the issue of reduced prediction accuracy arising from the redundant computation of dynamic parameter associations across multiple spatial dimensions resulting from the inherent token calculation strategy of Transformer based algorithms in short-term flight trajectory prediction, an in-depth analysis was conducted of the spatial-temporal characteristics of trajectory data and the computational logic of Transformer-like algorithms. Building upon a symmetrical multi-scale backbone architecture, a multi-dimensional splitting module, a local feature extraction module, and a global pattern fitting module were innovatively introduced, thereby a short-term flight trajectory prediction algorithm was proposed based on fusing global and local features in multiple dimensions. This algorithm mapped the short-term trajectory prediction task into a pattern regression problem involving an equal-length known trajectory sequence to an unknown sequence, effectively capturing the local spatial-temporal dynamic features and global spatial-temporal dependencies of the data in each dimension, and thereby further enhancing the model’s adaptability and robustness to complex trajectory patterns. Utilizing the massive real-world trajectory data from the OpenSky dataset, comparative experiments and ablation experiments were conducted. The results demonstrate that, compared with the latest Transformer-based baseline algorithm, this method achieves performance improvements of 10.92%, 14.39%, and 14.96% in terms of mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE), respectively. Furthermore, through module-by-module ablation verification, the accuracy of the algorithm gradually increases, fully demonstrating the effectiveness, advancement, and contribution of each module to the overall performance of the proposed algorithm.
SUI Jian-ming, TAO Ye, MA Ming-rong, YAN Zhen.
Short-term Prediction of Flight Trajectory Algorithm Based on Fusing Global and Local Features in Multiple Dimensions[J].
Science Technology and Engineering, 2026, 26(19): 8062-8073 DOI:10.12404/j.issn.1671-1815.2506309
在全球经济持续增长的背景下,航空运输需求呈现显著上升趋势,直接导致航班密度激增与空域复杂化程度加剧[1]。为应对这一挑战,交通预测技术已成为优化航班调度、提升民航运行效率的关键研究方向。其中,飞行航迹预测作为基于航迹运行(trajectory based operation,TBO)体系的关键支撑,受到欧洲单一天空空中交通管理研究计划(single European sky ATM research,SESAR)及美国下一代航空运输系统(NextGen)等重大战略项目的重点关注[2]。
在短期飞行轨迹预测领域,现有技术框架可大致分为空气动力学方法、状态估计方法、机器学习方法及深度学习方法四类[4]。其中,空气动力学方法主要依托预设规则与基于先验假设的数学模型,模拟航空器的近期飞行状态。波音欧洲研究公司与欧洲航行安全组织(European Organization for the Safety of Air Navigation,EUROCONTROL)合作,基于飞机性能数据库(base of aircraft data,BADA)开发了航迹预测算法[5],该算法融合多种天气条件,并将飞机意图抽象为指令信息,实现多航迹的动态计算。但此类方法需大量先验知识(如发动机参数等)支撑实际运算,计算效率较低,且在复杂多变的真实场景中获取足够参数的难度较大。
ChenZhijie, TangJinhui, WangChong, et al. Using artificial intelligence in airspace system to improve airspace hierachical governance capability[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 7-15.
[3]
ZhangZ, GuoD, ZhouS, et al. Flight trajectory prediction enabled by time-frequency wavelet transform[J]. Nature Communications, 2023, 14(1): 5258-5273.
[4]
GuoD, WuE, WuY, et al. FlightBERT: binary encoding representation for flight trajectory prediction[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24(2): 1828-1842.
CuiYaqi, XiongWei, HeYou. Adaptive forecast model for uncertain track[J]. Acta Aeronautica et Astronautica Sinica, 2019, 40(5): 241-250.
[7]
GalloE, López-LeonézJ, VilaplanaM, et al. Trajectory computation infrastructure based on BADA aircraft performance model[C]// Proceedings of the 26th Digital Avionics Systems Conference. Dallas: IEEE, 2007: 4-13.
ZhangZhaoning, WangJinghua. State space model of airport delay propagation under large-area flight delay[J]. Science Technology and Engineering, 2018, 18(31): 241-245.
ChenMingqiang, FuJiayun. Flight track prediction method based on unscented Kalman filter[J]. Computer Simulation, 2021, 38(6): 27-30, 6.
[12]
AlligierR, GianazzaD. Learning aircraft operational factors to improve aircraft climb prediction: a large scale multi-airport study[J]. Transportation Research Part C-Emerging Technologies, 2018, 96: 72-95.
ZhangYu, FengYahan, DingQianhui. Multi-sentiment classification of music lyrics by incorporating Word2Vec word embedding multi-core convolutional neural networks[J]. Science Technology and Engineering, 2024, 24(20): 8598-8605.
RenQingyang, WangYanding, ShiJian. Advances in target detection algorithms for convolutional neural networks[J]. Science Technology and Engineering, 2024, 24(32): 13665-13677.
[17]
ZhangW N, HuM H, DuJ H. An end-to-end framework for flight trajectory data analysis based on deep autoencoder network[J]. Aerospace Science and Technology, 2022, 127: 107726-107738.
[18]
ShiZ Y, XuM, PanQ. 4-D flight trajectory prediction with constrained LSTM network[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 22(11): 7242-7255.
[19]
WuX P, YangH Y, ChenH, et al. Long-term 4D trajectory prediction using generative adversarial networks[J]. Transportation Research Part C-Emerging Technologies, 2022, 136(13): 103554.
[20]
MaL, MengX R, WuZ J. Data-driven 4D trajectory prediction model using Attention-TCN-GRU[J]. Aerospace, 2024, 11(4): 21-43.
[21]
VaswaniA, ShazeerN, ParmarN, et al. Attention is all you need[C]// Proceedings of the 31st Annual Conference on Neural Information Processing Systems (NIPS). Long Beach: IEEE, 2017: 30-45.
[22]
TongQ, HuJ Q, ChenY L, et al. Long-term trajectory prediction model based on Transformer[J]. IEEE Access, 2023, 11: 143695-143703.
[23]
ZhangY H, YanJ C. Crossformer: transformer utilizing cross-dimension dependency for multivariate time series forecasting[C]// Proceedings of the the 11th International Conference on Learning Representations (ICLR). Kigila: AAAI, 2023: 1-21.
[24]
EkambaramV, JatiA, NguyenN, et al. TSMixer: lightweight MLP-mixer model for multivariate time series forecasting[C]// Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD). Long Beach: Assoc Computing Machinery, 2023: 459-4693.
LiuSifan, QinHuawang. Grid wind speed prediction based on DSS-UNet algorithm[J]. Science Technology and Engineering, 2024, 24(26): 11074-11080.
[27]
YeH R, XuD. InvPT++: inverted pyramid multi-task Transformer for visual scene understanding[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024, 46(12): 7493-7508.
[28]
ZhengX T, WangB Q, Du XiQ, et al. Mutual attention inception network for remote sensing visual question answering[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 14-28.
[29]
ShomanM, Amo-BoatengM, Adu-GyamfiY. Multi-purpose, multi-step deep learning framework for network-level traffic flow prediction[J]. Advances in Data Science and Adaptive Analysis, 2022, 14: 1-28.
[30]
TaoY, TangJ, ZhaoX, et al. Multi-scale network with attention mechanism for underwater image enhancement[J]. Neurocomputing, 2024, 595: 127926.
[31]
AlrahhalM, BaziY, AlhichriH, et al. Deep learning approach for active classification of electrocardiogram signals[J]. Information Sciences, 2016, 345: 340-354.
[32]
LaudaniA, LozitoG, FulgineiF, et al. On training efficiency and computational costs of a feed forward neural network: a review[J]. Computational Intelligence and Neuroscience, 2015, 2015: 818243.
ZhaoXin, LiXinjie, XuJian, et al. Parallel medical image registration model based on convolutional neural network and Transformer[J]. Journal of Computer Applications, 2024, 44(12): 3915-3921.
[35]
TaoY, ZhaoX, YanZ, et al. TiI: a novel framework for predicting pre-tactical multiple airports traffic flows by integrating flight plans and weather data[J]. IEEE Transactions on Intelligent Transportation Systems, 2025, 26: 12-24.
[36]
LiY, FangY, LongT. Noise robust aircraft trajectory prediction via autoregressive transformers with hybrid positional encoding[J]. Scientific Reports, 2025, 15(1): 1-19.
[37]
ChenY, SunJ, LinY, et al. Hybrid N-inception-LSTM-based aircraft coordinate prediction method for secure air traffic[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(3): 2773-2783.
[38]
LiJ, SongL, LiuC. The cubic trigonometric automatic interpolation spline[J]. IEEE-CAA Journal of Automatica Sinca, 2018, 5(6): 1136-1141.
[39]
ReyadM, SarhanA, ArafaM. A modified Adam algorithm for deep neural network optimization[J]. Neural Computing & Applications, 2023, 35(23): 17095-17112.