计算机科学 ›› 2021, Vol. 48 ›› Issue (12): 195-203.doi: 10.11896/jsjkx.210400022

• 数据库&大数据&数据科学 • 上一篇    下一篇

基于3D卷积和LSTM编码解码的出行需求预测

滕建, 滕飞, 李天瑞   

  1. 西南交通大学计算机与人工智能学院 成都611756
    综合交通大数据应用技术国家工程实验室 成都611756
  • 收稿日期:2021-04-01 修回日期:2021-07-17 出版日期:2021-12-15 发布日期:2021-11-26
  • 通讯作者: 李天瑞(trli@swjtu.edu.cn)
  • 作者简介:TJ950@my.swjtu.edu.cn
  • 基金资助:
    国家重点研发计划项目(2019YFB2101802);四川省重点研发项目(2021YFG0312)

Travel Demand Forecasting Based on 3D Convolution and LSTM Encoder-Decoder

TENG Jian, TENG Fei, LI Tian-rui   

  1. School of Computing and Artificial Intelligence,Southwest Jiaotong University,Chengdu 611756,China
    National Engineering Laboratory of Integrated Transportation Big Data Application Technology,Chengdu 611756,China
  • Received:2021-04-01 Revised:2021-07-17 Online:2021-12-15 Published:2021-11-26
  • About author:TENG Jian,born in 1996,postgraduate.His main research interests include big data and urban computing.
    LI Tian-rui,born in 1969,professor,is a fellow of IRSS and senior member of China Computer Federation,ACM and IEEE.His main research interests include big data,data mining,granular computing and rough sets.
  • Supported by:
    National Key R & D Program of China (2019YFB2101802) and Key R & D Project of Sichuan Province (2021YFG0312).

摘要: 可靠的区域出行需求预测能够为交通资源的调度和规划提供合理有效的建议。但是,出行预测是一个非常具有挑战性的问题,面临海量的时空大数据建模问题,如何有效地提取时空大数据中的空间特征和时间特征,成为当前城市计算的研究热点。文中提出了一种基于3D卷积和编码-解码注意力机制的需求预测模型(3D Convolution and Encoder-Decoder Attention Demand Forecasting,3D-EDADF),用于同时预测城市区域的出行需求流入量和流出量。3D-EDADF模型首先利用3D卷积来提取时空数据的时空相关性,然后使用LSTM编码解码来对时间依赖性进行捕获,并结合注意力机制来描述流入流出的差异性。3D-EDADF模型对临近依赖性、日常依赖性和周期依赖性这3种时间依赖特征进行混合建模,然后将它们的多维特征进行加权融合得到最终的预测结果。采用真实的出行需求数据集进行了大量的实验,结果表明,与基准模型相比,3D-EDADF模型的整体预测误差较低,具有较好的预测性能。

关键词: 出行需求预测, 时空大数据, 3D卷积, 编码-解码, 注意力机制

Abstract: Reliable regional travel demand forecasting can provide reasonable and effective suggestions for the scheduling and planning of traffic resources.However,travel forecasting is a very challenging problem,facing massive spatial-temporal big data modeling problem.And how to extract the spatial and temporal features of the data effectively has become a research hotspot of urban computing.This paper proposes a demand forecasting model based on 3D deconvolution and encoder-decoder attention mechanism (in short 3D-EDADF),which is used to predict the inflow and outflow of travel demand in urban areas at the same time.3D-EDADF model first uses 3D convolution to extract spatial-temporal correlation of data,then uses LSTM encoder-decoder to capture temporal dependence,and combines attention mechanism to describe the difference of inflow and outflow.3D-EDADF model conducts hybrid modeling on the three time-dependent features of closeness dependency,daily dependency and periodic dependency,and then weights and fuses their multi-dimensional features to obtain the final prediction result.The experiments are carried out by using real travel demand data sets.The results show that compared with baseline models,the 3D-EDADF model has the lowest overall prediction error and has better prediction performance.

Key words: Travel demand prediction, Spatial-temporal big data, 3D convolution, Encoder-decoder, Attention mechanism

中图分类号: 

  • TP181
[1]YAO H,WU F,KE J,et al.Deep multi-view spatial-temporal network for taxi demand prediction[C]//2018 32nd AAAI Conference on Artificial Intelligence,AAAI.2018:2588-2595.
[2]CHU K F,LAM A Y S,LI V O K.Travel demand prediction using deep multi-scale convolutional LSTM network[C]//2018 21st International Conference on Intelligent Transportation Systems (ITSC).IEEE,2018:1402-1407.
[3]MACIEJEWSKI M,BISCHOFF J,NAGEL K.An assignment-based approach to efficient real-time city-scale taxi dispatching[J].IEEE Intelligent Systems,2016,31(1):68-77.
[4]WANG D,YANG Y,NING S.Deepstcl:A deep spatio-temporal convlstm for travel demand prediction[C]//2018 International Joint Conference on Neural Networks(IJCNN).IEEE,2018:1-8.
[5]YANG G F,XU R,QIN M,et al.Short-term traffic volume forecasting based on ARMA and Kalman filter[J].Journal of Zhengzhou University(Engineering Edition),2017,38(2):36-40.
[6]KANCHYMALAY K,SALIM N,SUKPRASERT A,et al. Multivariate time series forecasting of crude palm oil price using machine learning techniques[C]//IOP Conference Series:Materials Science and Engineering.IOP Publishing,2017,226(1):012117.
[7]LU J Z,CHENG H.Short-term traffic flow forecast based on modified GA optimized BP neural network [J].Journal of Hefei University of Technology (Natural Science Edition),2015,38(1):127-131.
[8]TEDJOPURNOMO D A,BAO Z,ZHENG B,et al.A Survey on Modern Deep Neural Network for Traffic Prediction:Trends,Methods and Challenges[J/OL].IEEE Transactions on Know-ledge and Data Engineering,2020,https://ieeexplore.ieee.org/document/9112608.
[9]LECUN Y,BENGIO Y,HINTON G.Deep learning[J].Nature,2015,521(7553):436-444.
[10]KRIZHEVSKY A,SUTSKEVER I,HINTON G E.Imagenet classification with deep convolutional neural networks[J].Communications of the ACM,2017,60(6):84-90.
[11]KARPATHY A,FEI-FEI L.Deep visual-semantic alignments for generatingimage descriptions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2015:3128-3137.
[12]CHO K,VAN MERRIENBOER B,GULCEHRE C,et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[J].arXiv:1406.1078,2014.
[13]SCHIMBINSCHI F,NGUYEN X V,BAILEY J,et al.Traffic forecasting in complex urban networks:Leveraging big data and machine learning[C]//2015 IEEE International Conference on Big Data.IEEE,2015:1019-1024.
[14]HABTEMICHAEL F G,CETIN M.Short-term traffic flow rate forecasting based on identifying similar traffic patterns[J].Transportation Research Part C:Emerging Technologies,2016,66:61-78.
[15]QIAN X,UKKUSURI S V,YANG C,et al.A model for short-term taxi demand forecasting accounting for spatio-temporal correlations[R].Transportation Research Board 96th Annual Meeting,2017.
[16]SAADI I,WONG M,FAROOQ B,et al.An investigation into machine learning approaches for forecasting spatio-temporal demand in ride-hailing service[J].arXiv:1703.02433,2017.
[17]ZHANG X,HUANG C,XU Y,et al.Spatial-Temporal Convolutional Graph Attention Networks for Citywide Traffic Flow Forecasting[C]//Proceedings of the 29th ACM International Conference on Information & Knowledge Management.2020:1853-1862.
[18]CHEN X,XIE X,TENG D.Short-term Traffic Flow Prediction Based on ConvLSTM Model[C]//2020 IEEE 5th Information Technology and Mechatronics Engineering Conference (ITOEC).IEEE,2020:846-850.
[19]REN Y,ZHAO D,LUO D,et al.Global-Local Temporal Convolutional Network for Traffic Flow Prediction[J/OL].IEEE Transactions on Intelligent Transportation Systems,2020.https://ieeexplore.ieee.org/document/9216498.
[20]GUO S,LIN Y,LI S,et al.Deep spatial-temporal 3D convolutional neural networks for traffic data forecasting[J].IEEE Transactions on Intelligent Transportation Systems,2019,20(10):3913-3926.
[21]GAO K,LI D,CHEN L,et al.Incorporating intra-flow depen- dencies and inter-flow correlations for traffic matrix prediction[C]//2020 IEEE/ACM 28th International Symposium on Qua-lity of Service (IWQoS).IEEE,2020:1-10.
[22]DU S D,LI T R,YANG Y,et al.An Sequence-to-Sequence Spatial-Temporal Attention Learning Model for Urban Traffic Flow Prediction [J].Journal of Computer Research and Development,2020,57(8):1715-1728.
[23]CHEN Y,ZOU X,LI K,et al.Multiple local 3D CNNs for region-based prediction in smart cities[J].Information Sciences,2021,542:476-491.
[24]ZHOU Y,LI J,CHEN H,et al.A spatiotemporal hierarchical attention mechanism-based model for multi-stepstation-level crowd flow prediction[J].Information Sciences,2021,544:308-324.
[25]DAI G,MA C,XU X.Short-term traffic flow prediction method for urban road sections based on space-time analysis and GRU[J].IEEE Access,2019,7:143025-143035.
[26]ZHOU W,YANG Y,ZHANG Y,et al.Deep Flexible Structured Spatial-Temporal Model for Taxi Capacity Prediction[J].Knowledge-Based Systems,2020,205:106286.
[27]ZHANG J B,ZHENG Y,SUN J K,et al.Flow Prediction in Spatio-Temporal Networks Based on Multitask Deep Learning[J].IEEE Transactions on Knowledge and Data Engineering,2019,32 (3):468-478.
[28]XU L H,GUO Y T.Short-term Forecasting Model of Demand for Network Booking Taxi Based on GWO-LSTM [J].Automation and Instrumentation,2020,35(5):86-90.
[29]MADAN R,MANGIPUDI P S.Predicting computer network traffic:a time series forecasting approach using DWT,ARIMA and RNN[C]//2018 Eleventh International Conference on Contemporary Computing (IC3).IEEE,2018:1-5.
[30]WANG Y,ZHU S,LI C.Research on Multistep Time Series Prediction Based on LSTM[C]//2019 3rd International Confe-rence on Electronic Information Technology and Computer Engineering (EITCE).IEEE,2019:1155-1159.
[1] 叶中玉, 吴梦麟. 融合时序监督和注意力机制的脉络膜新生血管分割[J]. 计算机科学, 2021, 48(8): 118-124.
[2] 王雷全, 候文艳, 袁韶祖, 赵欣, 林瑶, 吴春雷. 利用全局与局部帧级特征进行基于共享注意力的视频问答[J]. 计算机科学, 2021, 48(8): 145-149.
[3] 张瑾, 段利国, 李爱萍, 郝晓燕. 基于注意力与门控机制相结合的细粒度情感分析[J]. 计算机科学, 2021, 48(8): 226-233.
[4] 王炽, 常俊. 基于3D卷积神经网络的CSI跨场景手势识别方法[J]. 计算机科学, 2021, 48(8): 322-327.
[5] 宋龙泽, 万怀宇, 郭晟楠, 林友芳. 面向出租车空载时间预测的多任务时空图卷积网络[J]. 计算机科学, 2021, 48(7): 112-117.
[6] 桑春艳, 胥文, 贾朝龙, 文俊浩. 社交网络中基于注意力机制的网络舆情事件演化趋势预测[J]. 计算机科学, 2021, 48(7): 118-123.
[7] 卿来云, 张建功, 苗军. 在线异常事件检测的时序建模[J]. 计算机科学, 2021, 48(7): 206-212.
[8] 徐少伟, 秦品乐, 曾建朝, 赵致楷, 高媛, 王丽芳. 基于多级特征和全局上下文的纵膈淋巴结分割算法[J]. 计算机科学, 2021, 48(6A): 95-100.
[9] 刘翔宇, 蹇木伟, 鲁祥伟, 何为凯, 李晓峰, 尹义龙. 基于眼动点视觉先验与边缘优化的显著性检测[J]. 计算机科学, 2021, 48(6A): 107-112.
[10] 冯姣, 陆昶谕. 基于残差注意力网络的跨媒体检索方法[J]. 计算机科学, 2021, 48(6A): 122-126.
[11] 潘芳, 张会兵, 董俊超, 首照宇. 基于高效Transformer的中文在线课程评论方面情感分析[J]. 计算机科学, 2021, 48(6A): 264-269.
[12] 石恒, 张玲. 基于生成对抗网络的图像阴影消除算法[J]. 计算机科学, 2021, 48(6): 145-152.
[13] 曾友渝, 谢强. 基于改进RNN和VAR的船舶设备故障预测方法[J]. 计算机科学, 2021, 48(6): 184-189.
[14] 程旭, 崔一平, 宋晨, 陈北京, 郑钰辉, 史金钢. 基于时空注意力机制的目标跟踪算法[J]. 计算机科学, 2021, 48(4): 123-129.
[15] 王习, 张凯, 李军辉, 孔芳, 张熠天. 联合自注意力和循环网络的图像标题生成[J]. 计算机科学, 2021, 48(4): 157-163.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 仲伟波,李忠梅,石婕,陈忠铭. 一种用于设施农业的ZigBee-WiFi网关研制[J]. 计算机科学, 2014, 41(Z6): 484 -486 .
[2] 杨洁,王国胤,李帅. 基于边界域的邻域知识距离度量模型[J]. 计算机科学, 2020, 47(3): 61 -66 .
[3] 裴嘉震, 徐曾春, 胡平. 融合视点机制与姿态估计的行人再识别方法[J]. 计算机科学, 2020, 47(6): 164 -169 .
[4] 潘孝勤, 芦天亮, 杜彦辉, 仝鑫. 基于深度学习的语音合成与转换技术综述[J]. 计算机科学, 2021, 48(8): 200 -208 .
[5] 王俊, 王修来, 庞威, 赵鸿飞. 面向科技前瞻预测的大数据治理研究[J]. 计算机科学, 2021, 48(9): 36 -42 .
[6] 余力, 杜启翰, 岳博妍, 向君瑶, 徐冠宇, 冷友方. 基于强化学习的推荐研究综述[J]. 计算机科学, 2021, 48(10): 1 -18 .
[7] 王梓强, 胡晓光, 李晓筱, 杜卓群. 移动机器人全局路径规划算法综述[J]. 计算机科学, 2021, 48(10): 19 -29 .
[8] 高洪皓, 郑子彬, 殷昱煜, 丁勇. 区块链技术专题序言[J]. 计算机科学, 2021, 48(11): 1 -3 .
[9] 毛瀚宇, 聂铁铮, 申德荣, 于戈, 徐石成, 何光宇. 区块链即服务平台关键技术及发展综述[J]. 计算机科学, 2021, 48(11): 4 -11 .
[10] 陈先来, 赵晓宇, 曾工棉, 安莹. 基于区块链的患者在线交流模型[J]. 计算机科学, 2021, 48(11): 28 -35 .