|本期目录/Table of Contents|

[1]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.[doi:10.11731/j.issn.1673-193x.2025.05.023]
　ZHANG Libiao,WEN Xiangxi,WU Minggong,et al.LSTM-RF 4D trajectory prediction based on time series[J].Journal of Safety Science and Technology,2025,21(5):179-186.[doi:10.11731/j.issn.1673-193x.2025.05.023]
点击复制

基于时间序列的LSTM-RF 4D航迹预测*

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

卷:
21
期数:
2025年5期
页码:
179-186
栏目:
职业安全卫生管理与技术
出版日期:
2025-05-30

文章信息/Info

Title:
LSTM-RF 4D trajectory prediction based on time series
文章编号:
1673-193X(2025)-05-0179-08
作者:
张立彪; 温祥西; 吴明功; 王俊杰; 常万昇
（1.中国人民解放军空军工程大学 空管领航学院，陕西 西安710051；
2.中国人民解放军94188部队，陕西 西安 710077；
3.中国人民解放军93125部队，江苏 徐州 221000；
4.中国人民解放军空军工程大学 防空反导学院，陕西 西安 710051）
Author(s):
ZHANG Libiao;  WEN Xiangxi;  WU Minggong;  WANG Junjie;  CHANG Wansheng
(1.Air Traffic Control and Navigation School,Air Force Engineering University,Xi’an Shaanxi 710051,China;
2.Unit 94188 of the PLA,Xi’an Shaanxi 710077,China;
3.Unit 93125 of the PLA,Xuzhou Jiangsu 221000,China;
4.Air Defense and Antimissile School,Air Force Engineering University,Xi’an Shaanxi 710051,China)
关键词:
时间序列; 4D航迹预测; ADS-B数据; 深度学习
Keywords:
time series;  4D trajectory prediction;  ADS-B data;  deep learning
分类号:
X949
DOI:
10.11731/j.issn.1673-193x.2025.05.023
文献标志码:
A
摘要:
为应对民航运输量快速增长和空域资源日益紧张问题，空域管理的优化、飞行安全和效率的提升变得尤为重要，基于这一背景，提出1种结合长短时记忆网络（LSTM）与随机森林（random forest,RF）组合的4D航迹预测模型（LSTM-RF）。该模型通过真实ADS-B数据的仿真实验，采用经纬度特征，将长短时记忆网络和随机森林有效融合，利用LSTM提取时间序列中的短期和长期依赖特征，生成隐藏状态向量，通过RF对这些特征进行非线性回归建模，从而综合发挥LSTM的时序建模能力和RF的非线性处理能力。研究结果表明：LSTM-RF模型在预测精度和泛化能力方面显著优于传统RNN和LSTM模型，与单独使用RNN和LSTM相比，LSTM-RF模型在MSE,MAE和MAPE 3个评价指标下均表现最优。研究结果可为未来航迹预测提供理论参考和实践支持。
Abstract:
In response to the rapid growth of civil aviation traffic and the increasing tension in airspace resources,optimizing the airspace management and improving the flight safety and efficiency have become particularly important.Against this backdrop,a 4D trajectory prediction model (LSTM-RF) combining the long short-term memory (LSTM) and random forest (RF) was proposed.Through the simulation experiment of real ADS-B data,the model used the latitude and longitude features to effectively integrate LSTM with RF.LSTM was used to extract the short-term and long-term dependent features in time series,generate hidden state vectors,and perform nonlinear regression modeling on these features through RF,so as to comprehensively exert the time series modeling ability of LSTM and the nonlinear processing ability of RF.The results show that the LSTM-RF model significantly outperforms traditional RNN and LSTM models in terms of prediction accuracy and generalization ability.Compared to using RNN and LSTM alone,LSTM-RF model performs optimally under the three evaluation metrics of mean square error (MSE),mean absolute error (MAE),and mean absolute percentage error (MAPE).The research results can provide theoretical reference and practical support for future trajectory prediction.

参考文献/References:

［1］王超，郭九霞，沈志鹏.基于基本飞行模型的4D航迹预测方法［J］.西南交通大学学报，2009，44（2）：295-300. WANG Chao,GUO Jiuxia,SHEN Zhipeng.4D trajectory prediction method based on basic flight model［J］.Journal of Southwest Jiaotong University,2009,44(2):295-300.
［2］韩云祥，汤新民，韩松臣.基于混杂系统理论的无冲突4D航迹预测［J］.西南交通大学学报，2012，47（6）：1069-1074. HAN Yunxiang,TANG Xinmin,HAN Songchen.Conflict-free 4D trajectory prediction based on hybrid system theory［J］.Journal of Southwest Jiaotong University,2012,47(6):1069-1074.
［3］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.4D trajectory prediction and uncertainty analysis for departure aircraft［J］.Journal of Southwest Jiaotong University,2016,51(4):800-806.
［4］YUAN S J,XILING L,LIANG H.Research on method of trajectory prediction in aircraft flight Based on aircraft performance and historical track data［J］.Mathematical Problems in Engineering,2021,2021.
［5］吕波，王超.改进的扩展卡尔曼滤波在航空器4D航迹预测算法中的应用［J］.计算机应用,2021,41(增刊1):277-282. LYU Bo,WANG Chao.Application of improved extended kalman filter in aircraft 4D trajectory prediction algorithm［J］.Computer Applications,2021,41(Supplement 1):277-282.
［6］赵竞哲，田晗，赵利庆，等.基于非线性回归的目标航迹预测算法［J］.软件,2023,44(4):35-39. ZHAO Jingzhe,TIAN Han,ZHAO Liqing,et al.Target track prediction algorithm based on nonlinear regression ［J］.Software,2023,44(4):35-39.
［7］赵元棣，李科频，朱文心.航空器爬升与下降阶段4D航迹预测［J］.科学技术与工程,2023,23(17):7582-7588. ZHAO Yuandi,LI Kepin,ZHU Wenxin.4D trajectory prediction for aircraft climb and descent phases［J］.Science Technology and Engineering,2023,23(17):7582-7588.
［8］赵庆媛，叶春茂，鲁耀兵.基于随机森林的微动特征重要性评估研究［J］.现代防御技术,2022,50(4):124-131. ZHAO Qingyuan,YE Chunmao,LU Yaobing.Research on the importance evaluation of micromotion features based on random forest［J］.Modern Defense Technology,2022,50(4):124-131.
［9］董欣放.基于机器学习的终端区4D航迹预测研究［D］.南京：南京航空航天大学,2021.
［10］石庆研，王文青，韩萍.基于在线更新LSTM网络的短期4D航迹预测算法［J］.信号处理,2021,37(1):66-74. SHI Qingyan,WANG Wenqing,HAN Ping.Short-term 4D trajectory prediction algorithm based on online updated LSTM network［J］.Signal Processing,2021,37(1):66-74.
［11］陈凯达，朱永生，闫柯，等.基于LSTM的船舶航迹预测［J］.船海工程,2019,48(6):121-125. CHEN Kaida,ZHU Yongsheng,YAN Ke,et al.Ship track prediction based on LSTM［J］.Ship & Ocean Engineering,2019,48(6):121-125.
［12］刘龙庚，翟俐民，韩云祥.基于卷积LSTM模型的航空器轨迹预测［J］.计算机工程与设计,2022,43(4):1127-1133. LIU Longgeng,ZHAI Limin,HAN Yunxiang.Aircraft trajectory prediction based on convLSTM model ［J］.Computer Engineering and Design,2022,43(4):1127-1133.
［13］翟俐民.基于轨迹聚类和LSTM的航迹预测方法研究［D］.成都：四川大学,2021.
［14］韩萍，张启，石庆研，等.基于DBSCAN-GRU算法的终端区4D航迹预测［J］.信号处理,2023,39(3):439-449. HAN Ping,ZHANG Qi,SHI Qingyan,et al.4D trajectory prediction in terminal area based on DBSCAN-GRU algorithm［J］.Signal Processing,2023,39(3):439-449.
［15］翟文鹏，宋一峤，张兆宁.基于Transformer-GRU网络的4D航迹预测［J］.重庆交通大学学报,2024,43(6):94-101. ZHAI Wenpeng,SONG Yiqiao,ZHANG Zhaoning.4D trajectory prediction based on transformer-GRU network［J］.Journal of Chongqing Jiaotong University,2024,43(6):94-101.
［16］石庆研，赵婉彤，韩萍.基于时空特征提取的 GTA-Seq2Seq 多步 4D 航迹预测模型［J］.北京航空航天大学学报,2024,43(6):22-28. SHI Qingyan,ZHAO Wantong,HAN Ping.Multi-step 4D trajectory prediction model based on spatio-temporal feature extraction using GTA-Seq2Seq［J］.Journal of Beijing University of Aeronautics and Astronautics,2024,43(6):22-28.
［17］李华，张强，斯永坤.麻雀算法改进BP神经网络的4D航迹预测［J］.科学技术与工程,2024,24(31):13635-13641. LI Hua,ZHANG Qiang,SI Yongkun.4D trajectory prediction based on SSA optimizing BP neural network［J］.Science Technology and Engineering,2024,24(31):13635-13641.

相似文献/References:

[1]吴松林,杜 扬,李国庆,等.基于激波管的油气爆炸实验和数据分析[J].中国安全生产科学技术,2014,10(2):5.[doi:10.11731/j.issn.1673-193x.2014.02.001]
　WU Song lin,DU Yang,LI Guo qing,et al.Explosion experiment and data analysis of gasoline\|air mixture based on shock tube[J].Journal of Safety Science and Technology,2014,10(5):5.[doi:10.11731/j.issn.1673-193x.2014.02.001]
[2]王云海,李娟,李春民.尾矿坝浸润线数据挖掘预测模型的样本选取研究*[J].中国安全生产科学技术,2009,5(5):9.
　WANG Yun hai,LI Juan,LI Chun min.Research of selecting the training samples for the infiltration route prediction model in tailing[J].Journal of Safety Science and Technology,2009,5(5):9.
[3]彭东,罗周全,秦亚光,等.基于加权线性回归模型组的湖北省工矿事故死亡人数分析预测[J].中国安全生产科学技术,2015,11(11):167.[doi:10.11731/j.issn.1673-193x.2015.11.028]
　ENG Dong,LUO Zhou-quan,QIN Ya-guang,et al.Analysis and forecast of death toll for industrial and mining accidents in Hubei province based on weighted linear regression model group[J].Journal of Safety Science and Technology,2015,11(5):167.[doi:10.11731/j.issn.1673-193x.2015.11.028]
[4]李润求,吴莹莹,施式亮,等.煤矿瓦斯涌出时序预测的自组织数据挖掘方法[J].中国安全生产科学技术,2017,13(7):18.[doi:10.11731/j.issn.1673-193x.2017.07.003]
　LI Runqiu,WU Yingying,SHI Shiliang,et al.Research on self-organizing data mining method for time series prediction of gas emission in coal mine[J].Journal of Safety Science and Technology,2017,13(5):18.[doi:10.11731/j.issn.1673-193x.2017.07.003]
[5]赖祥威,夏云霓,郑万波,等.基于集成学习的改进灰色瓦斯浓度序列预测*[J].中国安全生产科学技术,2021,17(7):16.[doi:10.11731/j.issn.1673-193x.2021.07.003]
　LAI Xiangwei,XIA Yunni,ZHENG Wanbo,et al.Improved grey prediction of gas concentration sequence based on integrated learning[J].Journal of Safety Science and Technology,2021,17(5):16.[doi:10.11731/j.issn.1673-193x.2021.07.003]
[6]刘莹,杨超宇.基于多因素的LSTM瓦斯浓度预测模型*[J].中国安全生产科学技术,2022,18(1):108.[doi:10.11731/j.issn.1673-193x.2022.01.017]
　LIU Ying,YANG Chaoyu.LSTM gas concentration prediction model based on multiple factors[J].Journal of Safety Science and Technology,2022,18(5):108.[doi:10.11731/j.issn.1673-193x.2022.01.017]
[7]汪瑾,郭在军,业巧林,等.基于缓冲区重采样的LSTM林火预测模型*[J].中国安全生产科学技术,2023,19(2):195.[doi:10.11731/j.issn.1673-193x.2023.02.027]
　WANG Jin,GUO Zaijun,YE Qiaolin,et al.LSTM prediction model of forest fire based on buffer resampling[J].Journal of Safety Science and Technology,2023,19(5):195.[doi:10.11731/j.issn.1673-193x.2023.02.027]
[8]王世杰,肖萌,华国威,等.基于CNN-GRU的混凝土大坝扬压力预测模型研究*[J].中国安全生产科学技术,2023,19(7):99.[doi:10.11731/j.issn.1673-193x.2023.07.015]
　WANG Shijie,XIAO Meng,HUA Guowei,et al.Research on uplift pressure prediction of concrete dam based on CNN-GRU[J].Journal of Safety Science and Technology,2023,19(5):99.[doi:10.11731/j.issn.1673-193x.2023.07.015]
[9]由东冉,刘剑,曹鹏,等.基于深度学习的实时热压湿变通风阻力系数时序预测研究*[J].中国安全生产科学技术,2025,21(11):150.[doi:10.11731/j.issn.1673-193x.2025.11.018]
　YOU Dongran,LIU Jian,CAO Peng,et al.Research on time series prediction of ventilation resistance coefficient under real-time thermal-pressure-humidity variation based on deep learning[J].Journal of Safety Science and Technology,2025,21(5):150.[doi:10.11731/j.issn.1673-193x.2025.11.018]

备注/Memo

备注/Memo:
收稿日期： 2025-01-07
* 基金项目： 国家自然科学基金项目(71801221)；国家社会科学基金项目（22XGL001）
作者简介： 张立彪，硕士研究生，主要研究方向为交通运输。
通信作者： 吴明功，硕士，教授，主要研究方向为交通运输工程、航空管制指挥与安全。
更新日期/Last Update: 2025-05-26