|本期目录/Table of Contents|

[1]白鹏,吉乐乐,刘楠,等.基于Bi-LSTM-Transformer的终端区进场航班轨迹预测*[J].中国安全生产科学技术,2026,22(5):205-212.[doi:10.11731/j.issn.1673-193x.2026.05.025]
　Bai Peng,Ji Lele,Liu Nan,et al.Bi-LSTM-Transformer-based trajectory prediction for terminal area approach flights[J].Journal of Safety Science and Technology,2026,22(5):205-212.[doi:10.11731/j.issn.1673-193x.2026.05.025]
点击复制

基于Bi-LSTM-Transformer的终端区进场航班轨迹预测*

《中国安全生产科学技术》[ISSN:1673-193X/CN:11-5335/TB]

卷:
22
期数:
2026年5期
页码:
205-212
栏目:
公共安全与应急管理
出版日期:
2026-05-30

文章信息/Info

Title:
Bi-LSTM-Transformer-based trajectory prediction for terminal area approach flights
文章编号:
1673-193X(2026)-05-0205-08
作者:
白鹏; 吉乐乐; 刘楠; 张丁戈; 王旋
（1.中国民航大学 空中交通管理学院，天津 300300；
2.上海东方飞行培训有限公司，上海 200131；
3.天津航大数据有限公司，天津 300300）
Author(s):
Bai Peng;  Ji Lele;  Liu Nan;  Zhang Dingge;  Wang Xuan
(1.College of Air Traffic Management,Civil Aviation University of China,Tianjin 300300,China;
2.Shanghai Eastern Flight Training Co.,Ltd.,Shanghai 200131,China;
3.Tianjing Hangda Data Co.,Ltd.,Tianjin 300300,China)
关键词:
轨迹预测; 航空运输; 深度学习; Bi-LSTM-Transformer; 全局时间依赖
Keywords:
trajectory prediction;  air transportation;  deep learning;  Bi-LSTM-Transformer;  global temporal dependencie
分类号:
X949
DOI:
10.11731/j.issn.1673-193x.2026.05.025
文献标志码:
A
摘要:
为解决航空器轨迹预测中存在的局部时序特征提取不足、捕捉全局时间依赖关系不充分、预测精度有限等问题，提出1种融合Bi-LSTM与Transformer的深度学习模型（Bi-LSTM-Transformer）。该模型利用Bi-LSTM捕捉轨迹数据的时间序列依赖关系，并引入Transformer的多头注意力机制以捕捉全局时间依赖关系，并基于真实ADS-B轨迹数据进行对比研究。研究结果表明：该模型在RMSE（均方根误差）、MAE（平均绝对误差）和MAPE（平均绝对百分比误差）3项指标上相较于BiGRU、Bi-LSTM和Transformer模型分别最大降低了60.64%，59.58%和67.28%，验证了模型在轨迹预测任务中有着优秀的预测性能。研究结果可为基于轨迹运行下的航班轨迹预测与空管智能化决策提供技术参考。
Abstract:
In order to address the issues of insufficient extraction of local temporal features and global temporal dependencies and limited prediction accuracy in current aircraft trajectory prediction,this paper proposes a deep learning model named Bi-LSTM-Transformer that integrates Bidirectional Long Short-Term Memory (Bi-LSTM) and Transformer.The model utilizes Bi-LSTM to capture temporal sequential dependencies in trajectory data and incorporates the multi-head attention mechanism of the Transformer to capture global temporal dependencie.Experiments based on real automatic dependent surveillance-broadcast (ADS-B) trajectory data demonstrate that the proposed model achieves peak improvements of 60.64%,59.58%,and 67.28% in root mean square error (RMSE),mean absolute error (MAE),and mean absolute percentage error (MAPE),respectively,when compared to the bidirectional gated recurrent unit (Bi-GRU),Bi-LSTM,and Transformer models.These results validate the effectiveness and superiority of the Bi-LSTM-Transformer model in trajectory prediction tasks.The research outcomes can provide important technical support for flight trajectory prediction and intelligent decision-making in air traffic management under trajectory-based operations.

参考文献/References:

［1］FAA/Eurocontrol.Common trajectory prediction-related terminology［R］.USA:FAA/Eurocontrol,2004.
［2］徐正凤,曾维理,羊钊.航空器轨迹预测技术研究综述［J］.计算机工程与应用,2021,57(12):65-74. Xu Zhengfeng,Zeng Weili,Yang Zhao.A survey of aircraft trajectory prediction technology［J］.Computer Engineering and Applications,2021,57(12):65-74.
［3］汤新民,李腾,陈强超,等.基于交互式多模型的短期4D航迹预测［J］.武汉理工大学学报（交通科学与工程版）,2020,44(1):39-45. Tang Xinmin,Li Teng,Chen Qiangchao,et al.Short-term 4D trajectory prediction based on interactive multiple model［J］.Journal of Wuhan University of Technology(Transportation Science & Engineering),2020,44(1):39-45.
［4］Choi H,Deng Chuhao,Hwang I.Hybrid machine learning and estimation-based flight trajectory prediction in terminal airspace［J］.IEEE Access,2021,9:151186-151197.
［5］张军峰,葛腾腾,陈强,等.离场航空器四维航迹预测及不确定性分析［J］.西南交通大学学报,2016,51(4):800-806. Zhang Junfeng,Ge Tengteng,Chen Qiang,et al.Four-dimensional trajectory prediction and uncertainty analysis for departure aircraft［J］.Journal of Southwest Jiaotong University,2016,51(4):800-806.
［6］Lee J,Lee S,Hwang I.Hybrid system modeling and estimation for arrival time prediction in terminal airspace［J］.Journal of Guidance,Control,and Dynamics,2016,39(4):903-910.
［7］Xu Zhengfeng,Zeng Weili,Chu Xiao,et al.Multi-aircraft trajectory collaborative prediction based on social long short-term memory network［J］.Aerospace,2021,8(4):115.
［8］Tan Minghu,Shen Hong,Xi Kang,et al.Trajectory prediction of flying vehicles based on deep learning methods［J］.Applied Intelligence,2023,53(11):13621-13642.
［9］Sahadevan D,Ponnusamy P,Gopi V,et al.Ground-based 4d trajectory prediction using bi-directional LSTM networks［J］.Applied Intelligence,2022,52(14):16417-16434.
［10］Ma Lan,Tian Shan.A hybrid CNN-LSTM model for aircraft 4D trajectory prediction［J］.IEEE Access,2020,8:134668-134680.
［11］Jia Peiyan,Chen Huiping,Zhang Lei,et al.Attention-LSTM based prediction model for aircraft 4-D trajectory［J］.Scientific Reports,2022,12(1):1-11.
［12］孔建国,李龙超,梁海军,等.一种基于Transformer-TCN-GRU的未来空管4D航迹预测方法［J/OL］.电讯技术,1-19［2025-07-05］.https://doi.org/10.20079/j.issn.1001-893x.241111001. Kong Jianguo,Li Longchao,Liang Haijun,et al.A future ATC 4D trajectory prediction method based on Transformer-TCN-GRU［J/OL］.Telecommunication Engineering,1-19［2025-07-05］.https://doi.org/10.20079/j.issn.1001-893x.241111001.
［13］Tong Qiang,Hu Jinqing,Chen Yuli,et al.Long-term trajectory prediction model based on transformer［J］.IEEE Access,2023,11:143695-143703.
［14］张立彪,温祥西,吴明功,等.基于时间序列的LSTM-RF 4D航迹预测［J］.中国安全生产科学技术,2025,21(5):179-186. Zhang Libiao,Wen Xiangxi,Wu Minggong,et al.4D trajectory prediction based on time series LSTM-RF［J］.Journal of Safety Science and Technology,2025,21(5):179-186.
［15］焦卫东,杨蓓.MDAT-Net:一种融合MSC和时空注意力的TCN航迹预测方法［J/OL］.北京航空航天大学学报,1-15［2025-05-07］.https://doi.org/10.13700/j.bh.1001-5965.2023.0717. Jiao Weidong,Yang Bei.MDAT-Net:a trajectory prediction method integrating MSC and spatio-temporal attention with TCN［J/OL］.Journal of Beijing University of Aeronautics and Astronautics,1-15［2025-05-07］.https://doi.org/10.13700/j.bh.1001-5965.2023.0717.
［16］Dong Xingchen,Tian Yong,Dai Linyanran,et al.A new accurate aircraft trajectory prediction in terminal airspace based on spatio-temporal attention mechanism［J］.Aerospace,2024,11(9):718.
［17］彭锐晖,贺基贤,孙殿星,等.复杂环境下基于CNN-Transformer的端到端航迹关联［J］.电讯技术,2025,65(5):663-673. Peng Ruihui,He Jixian,Sun Dianxing,et al.End-to-end track correlation based on CNN-Transformer in complex environment［J］.Telecommunication Engineering,2025,65(5):663-673.

相似文献/References:

[1]岳仁田,李君尉.基于多因素耦合的航空运输系统脆弱性分析*[J].中国安全生产科学技术,2020,16(12):150.[doi:10.11731/j.issn.1673-193x.2020.12.024]
　YUE Rentian,LI Junwei.Vulnerability analysis of air transportation system based on multifactor coupling[J].Journal of Safety Science and Technology,2020,16(5):150.[doi:10.11731/j.issn.1673-193x.2020.12.024]
[2]孙禾,陈一新.基于多标准和改进Siamese网络的相似航班号判断方法研究*[J].中国安全生产科学技术,2023,19(1):47.[doi:10.11731/j.issn.1673-193x.2023.01.007]
　SUN He,CHEN Yixin.Research on judgment method of similar call signs based on multi standards and improved Siamese network[J].Journal of Safety Science and Technology,2023,19(5):47.[doi:10.11731/j.issn.1673-193x.2023.01.007]
[3]刘颖俪,胡明华,苏佳明,等.基于多目标优化的机场机坪管制区域划分方法研究*[J].中国安全生产科学技术,2023,19(8):150.[doi:10.11731/j.issn.1673-193x.2023.08.022]
　LIU Yingli,HU Minghua,SU Jiaming,et al.Division method of airport apron control areas based on multi-objective optimization[J].Journal of Safety Science and Technology,2023,19(5):150.[doi:10.11731/j.issn.1673-193x.2023.08.022]
[4]张洪海,沈雪,瞿昕宜,等.基于BIRCH聚类的航空器自主运行风险状态评价*[J].中国安全生产科学技术,2024,20(9):188.[doi:10.11731/j.issn.1673-193x.2024.09.023]
　ZHANG Honghai,SHEN Xue,QU Xinyi,et al.Risk state evaluation on autonomous operation of aircraft based on BIRCH clustering[J].Journal of Safety Science and Technology,2024,20(5):188.[doi:10.11731/j.issn.1673-193x.2024.09.023]
[5]苏佳明,胡明华,尹嘉男,等.不同雪情条件下离场航班除冰与滑行协同调度方法研究*[J].中国安全生产科学技术,2025,21(12):177.[doi:10.11731/j.issn.1673-193x.2025.12.022]
　SU Jiaming,HU Minghua,YIN Jianan,et al.Research on coordinated scheduling methods for deicing and taxiing of departure flights under different snow conditions[J].Journal of Safety Science and Technology,2025,21(5):177.[doi:10.11731/j.issn.1673-193x.2025.12.022]
[6]罗诗涵,张晨阳,朱文兵,等.基于非接触式感知的飞行学员认知负荷评估研究*[J].中国安全生产科学技术,2025,21(2):67.[doi:10.11731/j.issn.1673-193x.2025.02.009]
　LUO Shihan,ZHANG Chenyang,ZHU Wenbing,et al.Study on cognitive load assessment of flight students based on non-contact perception[J].Journal of Safety Science and Technology,2025,21(5):67.[doi:10.11731/j.issn.1673-193x.2025.02.009]

备注/Memo

备注/Memo:
收稿日期： 2025-09-24；修回日期：2026-03-26
* 基金项目： 国家重点研发计划项目(2023YFB4302903);中国民用航空局安全能力建设项目（FSSA0005）;天津市科技局科学技术普及项目(24KPHDRC00060)
作者简介： 白鹏，硕士，高级实验师，主要研究方向为交通运输，人工智能等。
更新日期/Last Update: 2026-06-03