Flight Delay Propagation Prediction Model Based on CBAM-CondenseNet
-
摘要:
针对航班延误衍生的航班延误波及问题，该文提出一种基于CBAM-CondenseNet的航班延误波及预测模型。首先，通过分析航班延误在航空网络内产生的延误波及现象，确定会受前序延误航班影响的航班链；其次，对选定的航班链数据进行清洗，将航班信息与机场信息进行数据融合；最后，提出改进的CBAM-CondenseNet算法对融合后的数据进行特征提取，构建Softmax分类器对首班离港航班延误波及的后续离港航班延误等级进行预测。该文提出的CBAM-CondenseNet算法融合了CondenseNet和CBAM的优势，采用通道和空间注意力机制来加强网络结构深层信息的传递。实验结果表明，算法改进后有效提升网络性能，预测准确率可达97.55%。
-
关键词:
- 航班延误波及预测 /
- CBAM-CondenseNet /
- 数据融合 /
- 注意力机制
Abstract:For the problem of flight delay propagation caused by flight delay, a flight delay wave prediction model based on CBAM-CondenseNet is presented. Firstly, by analyzing the delays propagation in the aviation network caused by flight delays, the flight chain affected by the pre-order delays is determined; Secondly, the selected flight chain data is cleaned and the flight information and airport information are fused; Finally, an improved CBAM-CondenseNet algorithm is proposed to extract the number of fused flights. According to feature extraction, a Softmax classifier is constructed to predict the delays of the first departure flights and the subsequent flights. The CBAM-CondenseNet algorithm proposed in this paper combines the advantages of CondenseNet and CBAM, and uses channel and spatial attention mechanism to enhance the transmission of deep information in network structure. The experimental results show that the improved algorithm can effectively improve the network performance, and the prediction accuracy can reach 97.55%.
-
表 1 航班延误等级划分
延误等级 延误时间 未延误 $T \leq 15$ 轻度延误 $15 < T \leq 60$ 中度延误 $60 < T \leq 120$ 高度延误 $120 < T \leq 240$ 重度延误 $T > 240$ 
下载: 导出CSV
表 2 算法改进前后FLOPs(M)
网络层数 CondenseNet CBAM-CondenseNet 18 20.55 20.77 36 165.45 165.94 44 218.27 218.87
下载: 导出CSV
表 4 分类准确率对比(%)
网络层数 CondenseNet CBAM-CondenseNet 18 93.77 94.19 28 95.96 96.96 36 96.52 97.45 44 96.66 97.55 70 96.56 97.56 102 96.66 97.55 126 96.75 97.55
下载: 导出CSV
表 5 不同算法模型分类准确率对比(%)
网络
层数DenseNet SE-
DenseNetCondenseNet CBAM-
CondenseNet18 91.86 92.33 93.77 94.19 36 92.28 92.80 96.52 97.45 44 92.57 93.14 96.66 97.55
下载: 导出CSV
-
丁建立, 陈坦坦, 刘玉洁. 有色-时间Petri网航班延误模型与波及分析[J]. 计算机集成制造系统, 2008, 14(12): 2334–2340. doi: 10.13196/j.cims.2008.12.48.dingjl.001DING Jianli, CHEN Tantan, and LIU Yujie. Colored-timed Petri nets model of flight delays and propagated analysis[J]. Computer Integrated Manufacturing Systems, 2008, 14(12): 2334–2340. doi: 10.13196/j.cims.2008.12.48.dingjl.001 刘玉洁, 何丕廉, 刘春波, 等. 基于贝叶斯网络的航班延误波及研究[J]. 计算机工程与应用, 2006, 44(17): 242–245. doi: 10.3778/j.issn.1002-8331.2008.17.072LIU Yujie, HE Pilian, LIU Chunbo, et al. Flight delay propagation research based on Bayesian net[J]. Computer Engineering and Applications, 2006, 44(17): 242–245. doi: 10.3778/j.issn.1002-8331.2008.17.072 PYRGIOTIS N, MALONE K M, and ODONI A. Modelling delay propagation within an airport network[J]. Transportation Research Part C: Emerging Technologies, 2013, 27: 60–75. doi: 10.1016/j.trc.2011.05.017 邵荃, 朱燕, 贾萌, 等. 基于复杂网络理论的航班延误波及分析[J]. 航空计算技术, 2015, 45(4): 24–28. doi: 10.3969/j.issn.1671-654X.2015.04.007SHAO Quan, ZHU Yan, JIA Meng, et al. Analysis of flight delay propagation based on complex network theory[J]. Aeronautical Computing Technique, 2015, 45(4): 24–28. doi: 10.3969/j.issn.1671-654X.2015.04.007 CAMPANELLI B, FLEURQUIN P, ARRANZ A, et al. Comparing the modeling of delay propagation in the US and European air traffic networks[J]. Journal of Air Transport Management, 2016, 56: 12–18. doi: 10.1016/j.jairtraman.2016.03.017 WU Weiwei and WU C L. Enhanced delay propagation tree model with Bayesian network for modelling flight delay propagation[J]. Transportation Planning and Technology, 2018, 41(3): 319–335. doi: 10.1080/03081060.2018.1435453 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. doi: 10.1016/j.ifacol.2016.07.060 KHANMOHAMMADI S, TUTUN S, and 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. doi: 10.1016/j.procs.2016.09.321 TAKEICHI N, KAIDA R, SHIMOMURA A, et al. Prediction of delay due to air traffic control by machine learning[C]. AIAA Modeling and simulation Technologies Conference. Grapevine, USA, 2017: 191–199. HUANG Gao, LIU Shichen, VAN DER MAATEN L, et al. CondenseNet: An efficient DenseNet using learned group convolutions[C]. The IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA, 2018: 2752–2761. WOO S, PARK J, LEE J Y, et al. CBAM: Convolutional block attention module[C].The 15th European Conference on Computer Vision, Munich, Germany, 2018: 3–19. 盖杉, 鲍中运. 基于改进深度卷积神经网络的纸币识别研究[J]. 电子与信息学报, 2019, 41(8): 1992–2000. doi: 10.11999/JEIT181097GAI Shan and BAO Zhongyun. Banknote recognition research based on improved deep convolutional neural network[J]. Journal of Electronics &Information Technology, 2019, 41(8): 1992–2000. doi: 10.11999/JEIT181097 RUMELHART D E, HINTON G E, and WILLIAMS R J. Learning representations by back-propagating errors[J]. Nature, 1986, 323(6088): 533–536. doi: 10.1038/323533a0 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. doi: 10.11959/j.issn.1000-436x.2019091QU Jingyi, YE Meng, and QU Xing. Airport delay prediction model based on regional residual and LSTM network[J]. Journal on Communications, 2019, 40(4): 149–159. doi: 10.11959/j.issn.1000-436x.2019091 吴仁彪, 赵婷, 屈景怡. 基于深度SE-DenseNet的航班延误预测模型[J]. 电子与信息学报, 2019, 41(6): 1510–1517. doi: 10.11999/JEIT180644WU Renbiao, ZHAO Ting, and QU Jingyi. Flight delay prediction model based on deep SE-DenseNet[J]. Journal of Electronics &Information Technology, 2019, 41(6): 1510–1517. doi: 10.11999/JEIT180644 -
计量
- 文章访问数: 2637
- HTML全文浏览量: 1327
- PDF下载量: 160
- 被引次数: 0
下载:
