《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (10): 3292-3299.DOI: 10.11772/j.issn.1001-9081.2021081387

• 前沿与综合应用 • 上一篇    

基于注意力机制和生成对抗网络的飞行器短期航迹预测模型

陈玉立1,2, 佟强1,2, 谌彤童3, 侯守璐1, 刘秀磊1,2   

  1. 1.数据与科学情报分析研究所(北京信息科技大学), 北京 100101
    2.北京材料基因工程高精尖创新中心(北京信息科技大学), 北京 100101
    3.北京跟踪与通信技术研究所, 北京 100094
  • 收稿日期:2021-08-03 修回日期:2021-11-12 接受日期:2021-11-21 发布日期:2022-01-07 出版日期:2022-10-10
  • 通讯作者: 佟强
  • 作者简介:第一联系人:陈玉立(1994—),男,河南周口人,硕士研究生,主要研究方向:机器学习、数据挖掘、时间序列预测
    佟强(1985—),男(锡伯族),辽宁沈阳人,讲师,博士,CCF会员,主要研究方向:图像识别、计算机视觉、机器学习; tongq85@bistu.edu.cn
    谌彤童(1987—),男,江西南昌人,博士,主要研究方向:模式识别、实验鉴定
    侯守璐(1989—),女,河南驻马店人,讲师,博士,CCF会员,主要研究方向:物联网、边缘计算、绿色计算
    刘秀磊(1981—),男,河南濮阳人,副教授,博士,CCF会员,主要研究方向:语义Web、本体匹配、语义搜索、知识图谱、语义传感器。
  • 基金资助:
    北京信息科技大学校级基金资助项目(2121YJPY225);科研机构创新能力建设

Short-term trajectory prediction model of aircraft based on attention mechanism and generative adversarial network

Yuli CHEN1,2, Qiang TONG1,2, Tongtong CHEN3, Shoulu HOU1, Xiulei LIU1,2   

  1. 1.Institute of Data Science and Information Studies(Beijing Information Science and Technology University),Beijing 100101,China
    2.Beijing Advanced Innovation Center for Materials Genome Engineering;(Beijing Information Science and Technology University),Beijing 100101,China
    3.Beijing Institute of Tracking and Telecommunications Technology,Beijing 100094,China
  • Received:2021-08-03 Revised:2021-11-12 Accepted:2021-11-21 Online:2022-01-07 Published:2022-10-10
  • Contact: Qiang TONG
  • About author:CHEN Yul, born in 1994, M. S. candidate. His research interests include machine learning, data mining, time series prediction.
    TONG Qiang, born in 1985, Ph. D. , lecturer. His research interests include image recognition, computer vision, machine learning.
    CHEN Tongtong, born in 1987, Ph. D. . His research interests include pattern recognition, experimental identification
    HOU Shoulu, born in 1989, Ph. D. , lecturer. Her research interests include internet of things, edge computing, green computing.
    LIU Xiulei, born in 1981, Ph. D. , associate professor. His research interests include semantic Web, ontology matching, semantic search, knowledge graph, semantic sensor.
  • Supported by:
    Fund of Beijing Information Science and Technology University(2121YJPY225);Innovation Capacity Building of Scientific Research Institutions

摘要:

针对单一长短时记忆(LSTM)网络在航迹预测上无法有效提取关键信息以及难以精准拟合数据分布等问题,提出基于注意力机制和生成对抗网络(GAN)的飞行器短期轨迹预测模型。首先,引入注意力机制对航迹赋予不同的权重,以提升航迹中重要特征的影响力;其次,基于LSTM提取航迹序列特征,并经汇聚层汇集时间步长内所有的飞行器特征;最后,利用GAN在对抗博弈下不断优化的特性来优化模型,从而提高模型的准确性。相较于社会生成对抗网络(SGAN),所提模型在处于爬升阶段的数据集上的平均位移误差(ADE)、最终位移误差(FDE)及最大位移误差(MDE)分别降低了20.0%、20.4%和18.3%。实验结果表明,所提模型能更精确地预测未来航迹。

关键词: 航迹预测, 注意力机制, 生成对抗网络, 长短时记忆网络

Abstract:

Single Long Short-Term Memory (LSTM) network cannot effectively extract key information and cannot accurately fit data distribution in trajectory prediction. In order to solve the problems, a short-term trajectory prediction model of aircraft based on attention mechanism and Generative Adversarial Network (GAN) was proposed. Firstly, different weights were assigned to the trajectory by introducing attention mechanism, so that the influence of important features in the trajectory was able to be improved. Secondly, the trajectory sequence features were extracted by using LSTM, and the convergence net was used to gather all aircraft features within the time step. Finally, the characteristic of GAN optimizing continuously in adversarial game was used to optimize the model in order to improve the model accuracy. Compared with Social Generative Adversarial Network (SGAN), the proposed model has the Average Displacement Error (ADE), Final Displacement Error (FDE) and Maximum Displacement Error (MDE) reduced by 20.0%, 20.4% and 18.3% respectively on the dataset during climb phase. Experimental results show that the proposed model can predict future trajectories more accurately.

Key words: trajectory prediction, attention mechanism, Generative Adversarial Network (GAN), Long Short-Term Memory (LSTM) network

中图分类号: