基于区域残差和LSTM网络的机场延误预测模型

doi:10.11959/j.issn.1000-436x.2019091

摘要/Abstract

摘要：

针对目前民航运输业对机场延误预测高精度的要求，提出一种基于区域残差和长短时记忆（RR-LSTM）网络的机场延误预测模型。首先，将机场的属性信息、气象信息和相关运行航班信息进行融合；然后，利用RR-LSTM 网络对融合后的机场数据集进行特征提取；最后，构建 Softmax 分类器对机场延误分类预测。所提RR-LSTM网络模型既能有效提取机场延误数据的时间相关性，又能避免深层LSTM网络的梯度消失问题。实验结果表明，RR-LSTM网络模型预测准确率可达95.52%，取得了比传统网络模型更好的预测效果。其中，融合机场的气象信息和相关运行航班信息后，预测准确率可提高约11%。

关键词: 区域残差网络, 长短时记忆网络, 机场延误预测, 特征提取

Abstract:

Nowadays,the civil aviation industry has a high precision requirement of airport delay prediction,so an airport delay prediction model based on the RR-LSTM network was proposed.Firstly,the airport information,meteorological information and related flight information were integrated.Then,the RR-LSTM network was used to extract the features of the fused airport data set.Finally,the Softmax classifier was adopted to classify and predict the airport delay.The proposed RR-LSTM network model can not only extract the time correlation of airport delay data effectively,but also avoid the gradient disappearance problem of deep LSTM network.The experimental results indicate that the RR-LSTM network model has a prediction accuracy of 95.52%,which achieves better prediction results than the traditional network model.The prediction accuracy can be improved about 11% by fusing the weather information and the flight information of the airport.

Key words: regional residual network, long short term memory network, airport delay prediction, feature extraction

中图分类号:

TP183

屈景怡,叶萌,渠星. 基于区域残差和LSTM网络的机场延误预测模型[J]. 通信学报, 2019, 40(4): 149-159.

Jingyi QU,Meng YE,Xing QU. Airport delay prediction model based on regional residual and LSTM network[J]. Journal on Communications, 2019, 40(4): 149-159.

图/表 14

图1

图2

图3

图4

图5

表1

表2

表3

表4

图6

图7

表5

图8

表6

参考文献 26

[1]	吴薇薇, 孟亭婷, 张皓瑜 . 基于机场延误预测的航班计划优化研究[J]. 交通运输系统工程与信息, 2016,16(6): 189-195.
	WU W W , MENG T T , ZHANG H Y . Flight plan optimization based on airport delay prediction[J]. Journal of Transportation Systems En-gineering and Information Technology, 2016,16(6): 189-195.
[2]	NAYAK N , ZHANG Y . Estimation and comparison of impact of single airport delay on national airspace system with multivariate simultaneous models[J]. Transportation Research Record, 2011,2206(1): 52-60.
[3]	EADS G , KIEFER M , MEHNDIRATTA S . Short-term delay mitigation strategies for san francisco international airport[J]. Transportation Research Record Journal of the Transportation Research Board, 2001,1744(1): 44-51.
[4]	韩淑敏 . 机场运行可预测性分析与优化[D]. 天津:中国民航大学, 2016.
	HAN S M . Analysis and optimization of the airport operational pre-dictability[D]. Tianjing:Civil Aviation University of China, 2016.
[5]	罗谦, 张永辉, 程华 ,等. 基于航空信息网络的枢纽机场航班延误预测模型[J]. 系统工程理论与实践, 2014,34(S1): 143-150.
	LUO Q , ZHANG Y H , CHENG H ,et al. Study on flight delay predic-tion model based on flight networks[J]. Systems Engineer-ing-Theory＆Practice, 2014,34(S1): 143-150.
[6]	DEUTSCHMANN A , . Prediction of airport delays based on non-linear considerations of airport systems[C]// The 28th International Congress of the Aeronautical Sciences. 2012: 1-5.
[7]	徐涛, 丁建立, 顾彬 ,等. 基于增量式排列支持向量机的机场航班延误预警[J]. 航空学报, 2009,30(7): 1256-1263.
	XU T , DING J L , GU B ,et al. Forecast warning level of flight delays based on incremental ranking support vector machine[J]. Acta Aeronautica et Astronautica Sinica, 2009,30(7): 1256-1263.
[8]	张静, 徐肖豪, 王飞 ,等. 基于模糊线性回归模型的机场延误性能评估[J]. 交通运输工程学报, 2010,10(4): 109-114.
	ZHANG J , XU X H , WANG F ,et al. Airport delay performance evalu-ation based on fuzzy linear regression model[J]. Journal of Traffic and Transportation Engineering, 2010,10(4): 109-114.
[9]	MUKHERJEE A , GRABBE S , SRIDHAR B . Predicting ground delay program at an airport based on meteorological conditions[C]// The 14th AIAA Aviation Technology,Integration,and Operations Conference. 2014.
[10]	郭野晨风, 李杰, 胡明华 ,等. 基于简化WITI指标的机场延误预测方法[J]. 交通运输系统工程与信息, 2017,17(5): 207-213.
	GUO Y C F , LI J , HU M H ,et al. Airport delay prediction method based on simplified WITI index[J]. Journal of Transportation Systems Engineering and Information Technology, 2017,17(5): 207-213.
[11]	NOBORU T , RYOSUKE K , AKIHIDE S ,et al. Prediction of delay due to air traffic control by machine learning[C]// AIAA Modeling and Simulation Technologies Conference. 2017: 191-199
[12]	BASPINAR B , URE N K , KOYUNCU E ,et al. Analysis of delay characteristics of european air traffic through a data-driven airport-centric queuing network model[J]. IFAC-PapersOnLine, 2016,49(3): 359-364.
[13]	KIM Y J , CHOI S , BRICENO S ,et al. A deep learning approach to flight delay prediction[C]// The 35th Digital Avionics Systems Conference. 2016: 67-72.
[14]	KHANMOHAMMADI S , TUTUN S , KUCUK Y . A new multilevel input layer artificial neural network for predicting flight delays at JFK airport[J]. Procedia Computer Science, 2016,95: 237-244.
[15]	LECUN Y , BENGIO Y , HINTON G . Deep learning[J]. Nature, 2015,521(7553): 436-444.
[16]	TSOI C A , SHAOHUA T . Recurrent neural networks:a constructive algorithm,and its properties[J]. Neuro computing, 1997,15(3-4): 309-326.
[17]	HOCHREITER S , SCHMMIDHUBER J . Long short-term memory[J]. Neural Computation, 1997,9(8): 1735-1780.
[18]	HE K , ZHANG X , REN S ,et al. Deep residual learning for image recognition[C]// The 26th IEEE Conference on Computer Vision and Pattern Recognition. 2016: 770-778.
[19]	WANG J , YANG Y , MAO J ,et al. CNN-RNN:a unified framework for multi-label image classication[C]// The 29th IEEE Conference on Computer Vision and Pattern Recognition. 2016: 2285-2294.
[20]	IOFFE S , SZEGEDY C . Batch normalization:accelerating deep network training by reducing internal covariate shift[C]// The 32nd International Conference on Machine Learning. 2015: 448-456.
[21]	NAIR V , HINTON G E . Rectified linear units improve restricted boltzmann machines[C]// The 27th International Conference on Machine Learning. 2010: 807-814.
[22]	屈景怡, 朱威, 吴仁彪 . 基于衰减因子的双通道神经网络图像分类算法[J]. 系统工程与电子技术, 2017,39(6): 1391-1399.
	QU J Y , ZHU W , WU R B . Image classification for dual-channel neural networks based on attenuation factor[J]. Systems Engineering and Electronics, 2017,39(6): 1391-1399.
[23]	黄文坚, 唐源 . TensorFlow 实战[M]. 北京: 电子工业出版社, 2017.
	HUANG W J , TANG Y . TensorFlow practice[M]. Beijing: Publishing House of Electronics IndustryPress, 2017.
[24]	吴仁彪, 李佳怡, 屈景怡 . 基于双通道卷积神经网络的航班延误预测模型[J]. 计算机应用, 2018,38(7): 2100-2112.
	WU R B , LI J Y , QU J Y . Flight delay prediction based on du-al-channel convolutional neural networks[J]. Journal of Computer Ap-plications, 2018,38(7): 2100-2112.
[25]	MICCI-BARRECA D . A preprocessing scheme for high-cardinality categorical attributes in classification and prediction problems[J]. ACM SIGKDD Explorations Newsletter, 2001,3(1): 27-32.
[26]	DUAN K , KEERTHI S S , CHU W ,et al. Multi-category classification by soft-max combination of binary classifiers[C]// The International workshop on multiple classifier systems. 2003: 125-134.

网络名称		准确率
网络名称	数据集E		数据集E_t
ANN1	87.95%		89.49%
ANN2	88.89%		90.34%
ANN4	88.63%		90.15%
ANN6	87.02%		88.64%
VGG16	87.41%		88.93%
VGG19	88.58%		89.73%
ResNet18	90.05%		91.85%
ResNet34	92.14%		93.05%
ResNet50	92.37%		93.97%

不使用LSTM模块		使用LSTM模块
网络名称	准确率	网络名称	准确率
ANN1	89.49%	ANN1-LSTM	90.96%
ANN2	90.34%	ANN2-LSTM	91.01%
ANN4	90.15%	ANN4-LSTM	87.83%
ANN6	88.64%	ANN6-LSTM	84.52%
VGG16	88.93%	VGG16-LSTM	90.85%
VGG19	89.73%	VGG19-LSTM	91.24%
ResNet18	91.85%	RR18-LSTM	93.35%
ResNet34	93.05%	RR34-LSTM	94.96%
ResNet50	93.97%	RR50-LSTM	95.52%

网络名称	LSTM层数	全连接层数	网络总层数	准确率
LSTM-1	1	1	2	91.75%
LSTM-2	2	1	3	93.19%
LSTM-3	3	1	4	93.20%
LSTM-4	4	1	5	84.93%

网络名称	卷积层数	LSTM层数	全连接层数	网络总层数	准确率
RR18-LSTM	18	2	2	22	93.35%
RR34-LSTM	34	2	2	38	94.96%
RR50-LSTM	50	2	2	54	95.52%

网络名称	网络层数	测试准确率
网络名称	网络层数	维度重塑	单向池化
RR18-LSTM	22	92.82%	93.35%
RR34-LSTM	38	94.51%	94.96%
RR50-LSTM	54	95.06%	95.52%