《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (7): 2065-2072.DOI: 10.11772/j.issn.1001-9081.2023071045
收稿日期:
2023-08-02
修回日期:
2023-09-25
接受日期:
2023-10-09
发布日期:
2023-10-26
出版日期:
2024-07-10
通讯作者:
黄添强
作者简介:
李欢欢(1997—),女,贵州遵义人,硕士研究生,主要研究方向:深度学习、时空数据挖掘;基金资助:
Huanhuan LI, Tianqiang HUANG(), Xuemei DING, Haifeng LUO, Liqing HUANG
Received:
2023-08-02
Revised:
2023-09-25
Accepted:
2023-10-09
Online:
2023-10-26
Published:
2024-07-10
Contact:
Tianqiang HUANG
About author:
LI Huanhuan, born in 1997, M. S. candidate. Her research interests include deep learning, spatial-temporal data mining.Supported by:
摘要:
满足公众高质量出行需求是智能交通系统(ITS)的主要挑战之一。目前,针对公共交通出行需求预测问题,现有模型大多采用固定结构的图描述出行需求的空间相关性,忽略了出行需求在不同尺度下具有不同的空间依赖关系。针对上述问题,提出一种多尺度时空图卷积网络(MSTGCN)模型。该模型首先从全局尺度和局部尺度构建全局需求相似图和局部需求相似图,这2种图可以捕获公共交通出行需求长期内较为稳定的全局特征和短期内动态变化的局部特征。利用图卷积网络(GCN)提取2种图中的全局空间信息和局部空间信息,并引入注意力机制融合两种空间信息。为了拟合时间序列中潜藏的时间依赖关系,利用门控循环单元(GRU)捕捉公共交通需求的时变特征。采用纽约市出租车订单数据集和自行车订单数据集进行实验,结果表明MSTGCN模型在自行车订单数据集上均方根误差(RMSE)、平均绝对误差(MAE)和皮尔逊相关系数(PCC)达2.788 6、1.737 1、0.799 2,在出租车订单数据集上RMSE、MAE、PCC达9.573 4、5.861 2、0.963 1。可见,MSTGCN模型可以有效地挖掘公共交通出行需求的多尺度时空特性,对未来公共交通出行需求进行准确预测。
中图分类号:
李欢欢, 黄添强, 丁雪梅, 罗海峰, 黄丽清. 基于多尺度时空图卷积网络的交通出行需求预测[J]. 计算机应用, 2024, 44(7): 2065-2072.
Huanhuan LI, Tianqiang HUANG, Xuemei DING, Haifeng LUO, Liqing HUANG. Public traffic demand prediction based on multi-scale spatial-temporal graph convolutional network[J]. Journal of Computer Applications, 2024, 44(7): 2065-2072.
数据集 | 维度 | RMSE | MAE | PCC |
---|---|---|---|---|
自行车 | 32 | 3.216 5 | 1.932 0 | 0.720 3 |
64 | 2.967 9 | 1.828 4 | 0.768 0 | |
128 | 2.822 6 | 1.753 2 | 0.790 6 | |
256 | 2.788 6 | 1.737 1 | 0.799 2 | |
出租车 | 32 | 13.659 9 | 8.285 6 | 0.925 3 |
64 | 11.042 5 | 6.694 5 | 0.952 4 | |
128 | 9.573 4 | 5.861 2 | 0.963 1 | |
256 | 9.827 6 | 5.888 7 | 0.963 8 |
表1 MSTGCN模型设置不同隐藏层维度的预测效果
Tab. 1 Prediction results of MSTGCN model with different hidden layer dimensions
数据集 | 维度 | RMSE | MAE | PCC |
---|---|---|---|---|
自行车 | 32 | 3.216 5 | 1.932 0 | 0.720 3 |
64 | 2.967 9 | 1.828 4 | 0.768 0 | |
128 | 2.822 6 | 1.753 2 | 0.790 6 | |
256 | 2.788 6 | 1.737 1 | 0.799 2 | |
出租车 | 32 | 13.659 9 | 8.285 6 | 0.925 3 |
64 | 11.042 5 | 6.694 5 | 0.952 4 | |
128 | 9.573 4 | 5.861 2 | 0.963 1 | |
256 | 9.827 6 | 5.888 7 | 0.963 8 |
模型 | 自行车 | ||
---|---|---|---|
RMSE | MAE | PCC | |
HA | 5.200 3 | 3.461 7 | 0.166 9 |
XGBoost | 4.049 4 | 2.469 0 | 0.486 1 |
STGCN | 3.604 2 | 2.760 5 | 0.731 6 |
DCRNN | 3.209 4 | 1.895 4 | 0.722 7 |
STG2Seq | 3.984 3 | 2.497 6 | 0.515 2 |
Graph WaveNet | 3.294 3 | 1.991 1 | 0.700 3 |
CCRNN | |||
DG 2RNN | 2.960 1 | 1.816 2 | 0.775 5 |
本文模型 | 2.788 6 | 1.737 1 | 0.799 2 |
表2 不同模型在纽约自行车订单数据集上的预测结果
Tab. 2 Prediction results of different models on NYC Bike dataset
模型 | 自行车 | ||
---|---|---|---|
RMSE | MAE | PCC | |
HA | 5.200 3 | 3.461 7 | 0.166 9 |
XGBoost | 4.049 4 | 2.469 0 | 0.486 1 |
STGCN | 3.604 2 | 2.760 5 | 0.731 6 |
DCRNN | 3.209 4 | 1.895 4 | 0.722 7 |
STG2Seq | 3.984 3 | 2.497 6 | 0.515 2 |
Graph WaveNet | 3.294 3 | 1.991 1 | 0.700 3 |
CCRNN | |||
DG 2RNN | 2.960 1 | 1.816 2 | 0.775 5 |
本文模型 | 2.788 6 | 1.737 1 | 0.799 2 |
模型 | 出租车 | ||
---|---|---|---|
RMSE | MAE | PCC | |
HA | 29.780 6 | 16.150 9 | 0.633 9 |
XGBoost | 21.199 4 | 11.680 6 | 0.807 7 |
STGCN | 22.648 9 | 18.455 1 | 0.915 6 |
DCRNN | 14.792 6 | 8.427 4 | 0.912 2 |
STG2Seq | 18.045 0 | 9.941 5 | 0.865 0 |
GraphWaveNet | 13.072 9 | 8.103 7 | 0.932 2 |
CCRNN | |||
DG 2RNN | 9.359 0 | 5.379 9 | 0.966 6 |
本文模型 | 9.573 4 | 5.861 2 | 0.963 1 |
表3 不同模型在纽约出租车订单数据集上的预测结果
Tab. 3 Prediction results of different models on NYC Taxi dataset
模型 | 出租车 | ||
---|---|---|---|
RMSE | MAE | PCC | |
HA | 29.780 6 | 16.150 9 | 0.633 9 |
XGBoost | 21.199 4 | 11.680 6 | 0.807 7 |
STGCN | 22.648 9 | 18.455 1 | 0.915 6 |
DCRNN | 14.792 6 | 8.427 4 | 0.912 2 |
STG2Seq | 18.045 0 | 9.941 5 | 0.865 0 |
GraphWaveNet | 13.072 9 | 8.103 7 | 0.932 2 |
CCRNN | |||
DG 2RNN | 9.359 0 | 5.379 9 | 0.966 6 |
本文模型 | 9.573 4 | 5.861 2 | 0.963 1 |
数据集 | RMSE | MAE | PCC | |
---|---|---|---|---|
自行车 | 3 | 2.922 9 | 1.832 0 | 0.772 0 |
5 | 2.851 0 | 1.750 9 | 0.788 9 | |
7 | 2.788 6 | 1.737 1 | 0.799 2 | |
9 | 2.850 8 | 1.746 8 | 0.789 5 | |
出租车 | 3 | 18.154 6 | 10.895 0 | 0.865 1 |
5 | 9.573 4 | 5.861 2 | 0.963 1 | |
7 | 10.601 4 | 6.426 2 | 0.957 2 | |
9 | 15.689 6 | 9.929 6 | 0.902 9 |
表4 不同 k下的预测效果对比
Tab. 4 Prediction results with different k
数据集 | RMSE | MAE | PCC | |
---|---|---|---|---|
自行车 | 3 | 2.922 9 | 1.832 0 | 0.772 0 |
5 | 2.851 0 | 1.750 9 | 0.788 9 | |
7 | 2.788 6 | 1.737 1 | 0.799 2 | |
9 | 2.850 8 | 1.746 8 | 0.789 5 | |
出租车 | 3 | 18.154 6 | 10.895 0 | 0.865 1 |
5 | 9.573 4 | 5.861 2 | 0.963 1 | |
7 | 10.601 4 | 6.426 2 | 0.957 2 | |
9 | 15.689 6 | 9.929 6 | 0.902 9 |
数据集 | RMSE | MAE | PCC | |
---|---|---|---|---|
自行车 | 5 | 2.890 1 | 1.789 2 | 0.782 0 |
10 | 2.876 4 | 1.780 0 | 0.784 2 | |
15 | 2.788 6 | 1.737 1 | 0.799 2 | |
20 | 3.088 3 | 1.908 5 | 0.745 5 | |
出租车 | 5 | 9.573 4 | 5.861 2 | 0.963 1 |
10 | 10.322 8 | 6.184 0 | 0.958 4 | |
15 | 10.200 6 | 6.155 5 | 0.959 6 | |
20 | 24.508 1 | 14.633 0 | 0.733 0 |
表5 不同 β下的预测效果
Tab. 5 Prediction results with different β
数据集 | RMSE | MAE | PCC | |
---|---|---|---|---|
自行车 | 5 | 2.890 1 | 1.789 2 | 0.782 0 |
10 | 2.876 4 | 1.780 0 | 0.784 2 | |
15 | 2.788 6 | 1.737 1 | 0.799 2 | |
20 | 3.088 3 | 1.908 5 | 0.745 5 | |
出租车 | 5 | 9.573 4 | 5.861 2 | 0.963 1 |
10 | 10.322 8 | 6.184 0 | 0.958 4 | |
15 | 10.200 6 | 6.155 5 | 0.959 6 | |
20 | 24.508 1 | 14.633 0 | 0.733 0 |
1 | XU J, RAHMATIZADEH R, BÖLÖNI L, et al. Real-time prediction of taxi demand using recurrent neural networks [J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 19( 8): 2572- 2581. |
2 | ZHANG J, ZHENG Y, QI D. Deep spatio-temporal residual networks for citywide crowd flows prediction [C]// Proceedings of the 31 st AAAI Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2017: 1655- 1661. |
3 | YAO H, WU F, KE J, et al. Deep multi-view spatial-temporal network for taxi demand prediction [C]// Proceedings of the 32 nd AAAI Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2018: 2588- 2595. |
4 | LI Y, YU R, SHAHABI C, et al. Diffusion convolutional recurrent neural network: data-driven traffic forecasting [C/OL]// Proceedings of the 6th International Conference on Learning Representations. New York: PLMR [ 2023-07-01]. . |
5 | 周榆欣, 邬群勇. CLAB模型:一种乘客出租出行需求短时预测的深度学习模型[J]. 地球信息科学学报, 2023, 25( 1): 77- 89. |
ZHOU Y X, WU Q Y. CLAB model: a deep learning model for short-term prediction of passenger rental travel demand [J]. Journal of Geo-Information Science, 2023, 25( 1): 77- 89. | |
6 | 王博伟, 邓君, 吕斌. 基于WCGAN的出租车需求热点预测[J]. 计算机工程与应用, 2023, 59( 12): 293- 300. |
WANG B W, DENG J, LYU B. Hotspot forecast of taxi demand based on WCGAN [J]. Computer Engineering and Applications, 2023, 59( 12): 293- 300. | |
7 | 陈柘, 刘嘉华, 赵斌, 等. 基于GCN和TCN的多因素城市路网出租车需求预测[J]. 控制与决策, 2023, 38( 4): 1031- 1038. |
CHEN Z, LIU J H, ZHAO B, et al. Multi-factor taxi demand forecasting for urban road network based on GCN and TCN [J]. Control and Decision, 2023, 38( 4): 1031- 1038. | |
8 | 董成祥, 魏昕, 张坤鹏, 等. 基于图卷积网络的乘客打车需求预测[J]. 工业工程, 2022, 25( 5): 98- 105. |
DONG C X, WEI X, ZHANG K P, et al. Passenger ride-hailing demand prediction based on graph convolutional networks [J]. Industrial Engineering Journal, 2022, 25( 5): 98- 105. | |
9 | 魏远, 林彦, 郭晟楠, 等. 融合出发地与目的地时空相关性的城市区域间出租车需求预测[J]. 计算机应用, 2023, 43( 7): 2100- 2106. |
WEI Y, LIN Y, GUO S N, et al. Prediction of taxi demands between urban regions by fusing origin-destination spatial-temporal correlation [J]. Journal of Computer Applications, 2023, 43( 7): 2100- 2106. | |
10 | ZHANG D, LI J. Multi-view fusion neural network for traffic demand prediction [J]. Information Sciences, 2023, 646: 119303. |
11 | 黄昕, 毛政元. 基于时空多图卷积网络的网约车乘客需求预测[J]. 地球信息科学学报, 2023, 25( 2): 311- 323. |
HUANG X, MAO Z Y. Prediction of passenger demand for online car-hailing based on spatio-temporal multi-graph convolution network [J]. Journal of Geo-Information Science, 2023, 25( 2): 311- 323. | |
12 | XU H, ZOU T, LIU M, et al. Adaptive spatiotemporal dependence learning for multi-mode transportation demand prediction [J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23( 10): 18632- 18642. |
13 | SONG C, LIN Y, GUO S, et al. Spatial-temporal synchronous graph convolutional networks: a new framework for spatial-temporal network data forecasting [J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34( 1): 914- 921. |
14 | WU Z, PAN S, LONG G, et al. Graph wavenet for deep spatial-temporal graph modeling [C]// Proceedings of the 28th International Joint Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2019: 1907- 1913. |
15 | YE J, SUN L, DU B, et al. Coupled layer-wise graph convolution for transportation demand prediction [J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2021, 35( 5): 4617- 4625. |
16 | SHANG C, CHEN J, BI J. Discrete graph structure learning for forecasting multiple time series [EB/OL]. ( 2021-02-15) [ 2023-07-01]. . |
17 | YE J, LIU Z, DU B, et al. Learning the evolutionary and multi-scale graph structure for multivariate time series forecasting [C]// Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York: ACM, 2022: 2296- 2306. |
18 | MOREIRA-MATIAS L, GAMA J, FERREIRA M, et al. Predicting taxi-passenger demand using streaming data [J]. IEEE Transactions on Intelligent Transportation Systems, 2013, 14( 3): 1393- 1402. |
19 | TONG Y, CHEN Y, ZHOU Z, et al. The simpler the better: a unified approach to predicting original taxi demands based on large-scale online platforms [C]// Proceedings of the 23 rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2017: 1653- 1662. |
20 | WU C-H, J-M HO, LEE D-T. Travel-time prediction with support vector regression [J]. IEEE Transactions on Intelligent Transportation Systems, 2004, 5( 4): 276- 281. |
21 | GONG Y, ZHANG Y. Research of short-term traffic volume prediction based on Kalman filtering [C]// Proceedings of the 2013 6th International Conference on Intelligent Networks and Intelligent Systems. Piscataway: IEEE, 2013: 99- 102. |
22 | RODRIGUEZ A, LAIO A. Clustering by fast search and find of density peaks [J]. Science, 2014, 344( 6191): 1492- 1496. |
23 | YU B, YIN H, ZHU Z. Spatio-temporal graph convolutional networks: a deep learning framework for traffic forecasting [C]// Proceedings of the 27th International Joint Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2018: 3634- 3640. |
24 | BAI L, YAO L, KANHERE S S, et al. Stg2 seq: spatial-temporal graph to sequence model for multi-step passenger demand forecasting [C]// Proceedings of the 28th International Joint Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2019: 1981- 1987. |
25 | ZHAO J, CHEN C, LIAO C, et al. 2F-TP: learning flexible spatiotemporal dependency for flexible traffic prediction [J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24( 12): 15379- 15391. |
[1] | 潘烨新, 杨哲. 基于多级特征双向融合的小目标检测优化模型[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2871-2877. |
[2] | 赵志强, 马培红, 黑新宏. 基于双重注意力机制的人群计数方法[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2886-2892. |
[3] | 庞川林, 唐睿, 张睿智, 刘川, 刘佳, 岳士博. D2D通信系统中基于图卷积网络的分布式功率控制算法[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2855-2862. |
[4] | 秦璟, 秦志光, 李发礼, 彭悦恒. 基于概率稀疏自注意力神经网络的重性抑郁疾患诊断[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2970-2974. |
[5] | 王熙源, 张战成, 徐少康, 张宝成, 罗晓清, 胡伏原. 面向手术导航3D/2D配准的无监督跨域迁移网络[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2911-2918. |
[6] | 李力铤, 华蓓, 贺若舟, 徐况. 基于解耦注意力机制的多变量时序预测模型[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2732-2738. |
[7] | 薛桂香, 王辉, 周卫峰, 刘瑜, 李岩. 基于知识图谱和时空扩散图卷积网络的港口交通流量预测[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2952-2957. |
[8] | 李顺勇, 李师毅, 胥瑞, 赵兴旺. 基于自注意力融合的不完整多视图聚类算法[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2696-2703. |
[9] | 黄云川, 江永全, 黄骏涛, 杨燕. 基于元图同构网络的分子毒性预测[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2964-2969. |
[10] | 薛凯鹏, 徐涛, 廖春节. 融合自监督和多层交叉注意力的多模态情感分析网络[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2387-2392. |
[11] | 汪雨晴, 朱广丽, 段文杰, 李书羽, 周若彤. 基于交互注意力机制的心理咨询文本情感分类模型[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2393-2399. |
[12] | 高鹏淇, 黄鹤鸣, 樊永红. 融合坐标与多头注意力机制的交互语音情感识别[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2400-2406. |
[13] | 刘禹含, 吉根林, 张红苹. 基于骨架图与混合注意力的视频行人异常检测方法[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2551-2557. |
[14] | 李钟华, 白云起, 王雪津, 黄雷雷, 林初俊, 廖诗宇. 基于图像增强的低照度人脸检测[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2588-2594. |
[15] | 莫尚斌, 王文君, 董凌, 高盛祥, 余正涛. 基于多路信息聚合协同解码的单通道语音增强[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2611-2617. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||