计算机应用 ›› 2020, Vol. 40 ›› Issue (8): 2420-2427.DOI: 10.11772/j.issn.1001-9081.2019112061

• 应用前沿、交叉与综合 • 上一篇    下一篇

基于团簇随机连接的CliqueNet航班延误预测模型

屈景怡1, 曹磊1, 陈敏1, 董樑2, 曹烨琇2   

  1. 1. 天津市智能信号与图像处理重点实验室(中国民航大学), 天津 300300;
    2. 中国民用航空华东地区空中交通管理局, 上海 200335
  • 收稿日期:2019-12-05 修回日期:2020-03-12 出版日期:2020-08-10 发布日期:2020-06-29
  • 通讯作者: 屈景怡(1978-),女,天津人,副教授,博士,主要研究方向:航空运输大数据、深度学习,qujingyicauc@163.com
  • 作者简介:曹磊(1994-),男,安徽六安人,硕士研究生,主要研究方向:航空运输大数据、神经网络;陈敏(1987-),男,天津人,助理研究员,硕士,主要研究方向:信号仿真、雷达及ADS-B数据处理、空管自动化系统;董樑(1980-),男,上海人,工程师,硕士,主要研究方向:统计分析、数据挖掘、公共经济与管理;曹烨琇(1983-),女,上海人,工程师,主要研究方向:管制运行信息系统、大数据。
  • 基金资助:
    国家自然科学基金资助项目(U1833105);天津市自然科学基金资助项目(19JCYBJC15900)。

CliqueNet flight delay prediction model based on clique random connection

QU Jingyi1, CAO Lei1, CHEN Min1, DONG Liang2, CAO Yexiu2   

  1. 1. Tianjin Key Laboratory of Advanced Signal and Image Processing(Civil Aviation University of China), Tianjin 300300, China;
    2. CAAC East China Regional Administration, Shanghai 200335, China
  • Received:2019-12-05 Revised:2020-03-12 Online:2020-08-10 Published:2020-06-29
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (U1833105), the Natural Science Foundation of Tianjin (19JCYBJC15900).

摘要: 针对目前民航运输业延误率较高,而传统算法难以解决高精度延误预测的问题,提出一种基于随机连接团簇网络(CliqueNet)航班延误预测模型。该模型首先对航班数据和相关气象数据进行融合;然后,充分利用改进后的网络模型对融合后的数据集进行特征提取;最后,使用Softmax分类器进行航班离港延误各等级的高精度预测。模型的主要特点是:在团簇特征层的随机连接,以及在转换层引入通道和空间注意力残差(CSAR)模块。前者以更为有效的连接方式传递特征信息;后者则对特征信息进行通道和空间维度的双重标定,以提高准确率。实验结果表明,对融合数据进行预测,引入随机连接和CSAR模块后,新模型的准确率分别提高了0.5%、1.3%,最终准确率能达到93.40%。

关键词: 团簇网络, 随机连接, 特征重标定, 航班延误预测, 数据融合

Abstract: Aiming at the current high delay rate of the civil aviation transportation industry, and the fact that the high-precision delay prediction problem can hardly be solved by traditional algorithms, a randomly connected Clique Network (CliqueNet) based flight delay prediction model was proposed. Firstly, the flight data and related weather data were fused by the model. Then, making full use of the improved network model to extract features from the fused dataset. Finally, the softmax classifier was used to predict the flight departure delay of all levels with high precision. The main features of the model include random connection of clique feature layers and the introduction of Channel-wise and Spatial Attention Residual (CSAR) block to the transition layer. The former transmits the feature information in a more effective connection; and the latter double-calibrates the feature information on the channel and spatial dimensions to improve accuracy. Experimental results show that the prediction accuracy of the fused data is improved by 0.5% and 1.3% respectively with the introduction of random connection and CSAR block, and the final accuracy of the new model reaches 93.40%.

Key words: Clique Network (CliqueNet), random connection, feature recalibration, flight delay prediction, data fusion

中图分类号: