[1]贵向泉,熊家昌,李 立,等.基于 ASTG-CRNN 模型的多步长交通流预测[J].计算机技术与发展,2023,33(09):141-148.[doi:10. 3969 / j. issn. 1673-629X. 2023. 09. 021]
　GUI Xiang-quan,XIONG Jia-chang,LI Li,et al.Multi-step Traffic Flow Prediction Based on ASTG-CRNN[J].,2023,33(09):141-148.[doi:10. 3969 / j. issn. 1673-629X. 2023. 09. 021]
点击复制

基于 ASTG-CRNN 模型的多步长交通流预测()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
33
期数:
2023年09期
页码:
141-148
栏目:
人工智能
出版日期:
2023-09-10

文章信息/Info

Title:
Multi-step Traffic Flow Prediction Based on ASTG-CRNN
文章编号:
1673-629X(2023)09-0141-08
作者:
贵向泉1 ; 熊家昌1 ; 李 立1 ; 郭莎莎2
1. 兰州理工大学 计算机与通信学院,甘肃 兰州 730050;
2. 中国石油天然气股份有限公司长庆油田分公司 数字化与信息中心,陕西 西安 710000
Author(s):
GUI Xiang-quan1 ; XIONG Jia-chang1 ; LI Li1 ; GUO Sha-sha2
1. School of Computer and Communication,Lanzhou University of Technology,Lanzhou 730050,China;
2. Digitization and Information Center,China National Petroleum Corporation Changqing Oilfield Branch,Xi’an 710000,China
关键词:
交通流预测; 注意力机制; 相对邻近度; 时空相关性; 图卷积网络; 循环神经网络
Keywords:
traffic flow prediction; attention mechanism; relative proximity; spatio - temporal correlations; graph convolutional network; recurrent neural network
分类号:
TP183;U491. 1+4
DOI:
10. 3969 / j. issn. 1673-629X. 2023. 09. 021
摘要:
针对交通流预测模型中路网表征结构难以进行刻画和交通流数据中动态时空相关性难以进行建模以及其中时间特征捕获不充分的问题,提出一种基于注意力机制和时空图卷积循环神经网络的交通流预测模型( ASTG-CRNN) 。 首先,通过定义节点相对邻近度来确定路网表征结构的关系权重;其次,通过在时空维度上引入注意力机制对动态时空相关性进行建模,再采用图卷积捕获交通流数据中的空间特征;最后,采用卷积神经网络和双向门控循环神经单元的组合模块共同捕捉时间特征,从而能更好地表达交通流的时空特性。 在两个公开交通流数据集 PeMS04 和 PeMS08 上对模型预测效果进行验证,其结果表明,ASTG-CRNN 模型的预测结果均优于其它模型,与时空同步图卷积网络模型( STSGCN) 相比,在未来 1h 内预测结果的 MAE、RMSE 和 MAPE 在数据集 PeMS08 上分别降低了 2. 71、2. 69 和 0. 87% 。
Abstract:
Aiming at the problems that it is difficult to describe the road network representation structure in the traffic flow predictionmodel,to model the dynamic spatio-temporal correlation,and to capture sufficiently the temporal features in the traffic flow data,wepropose a traffic flow prediction model based on attention mechanism and spatio-temporal graph convolutional recurrent neural network( ASTG- CRNN) . Firstly, the relationship weight of the road network representation structure is determined by defining the relativeproximity of nodes. Secondly,the dynamic spatio-temporal correlation is modeled by introducing an attention mechanism in the spatio-temporal dimension,and then the spatial features in the traffic flow data are captured by the graph convolutional network. Finally,thecombination module of the convolutional neural network and the bidirectional gated recurrent neural unit are used to capture temporalfeatures,so that the spatio-temporal characteristics of traffic flow can be better expressed. The prediction effect of the model is verifiedon two public traffic flow datasets PeMS04 and PeMS08. It is showed that the prediction results of the ASTG-CRNN model are betterthan that of other models. Compared with the spatio-temporal synchronous graph convolutional network model ( STSGCN) ,the MAE,　RMSE,and MAPE of the predicting results in the next 1h are reduced by 2. 71,2. 69 and 0. 87% respectively on the dataset PeMS08.

相似文献/References:

[1]廖荣华,兰时勇,刘正熙. 基于混沌时间序列局域法的短时交通流预测[J].计算机技术与发展,2015,25(01):1.
　LIAO Rong-hua,LAN Shi-yong,LIU Zheng-xi. Short-term Traffic Flow Forecasting Based on Local Prediction Method in Chaotic Time Series[J].,2015,25(09):1.
[2]杨济瑞,赵海涛,刘南杰. 改进的三次指数平滑法及其在车联网中的应用[J].计算机技术与发展,2016,26(11):164.
　YANG Ji-rui,ZHAO Hai-tao,LIU Nan-jie. Modified Cubic Exponential Smoothing Algorithm and Its Application on IoV[J].,2016,26(09):164.
[3]成云,成孝刚,谈苗苗,等. 基于ARIMA和小波神经网络组合模型的交通流预测[J].计算机技术与发展,2017,27(01):169.
　CHENG Yun,CHENG Xiao-gang,TAN Miao-miao,et al. Traffic Flow Prediction Based on Hybrid Model of ARIMA and WNN[J].,2017,27(09):169.
[4]李梦洁,董峦.基于 PyTorch 的机器翻译算法的实现[J].计算机技术与发展,2018,28(10):160.[doi:10.3969/ j. issn.1673-629X.2018.10.033]
　LI Meng-jie,DONG Luan.Implementation of Machine Translation Algorithm Based on PyTorch[J].,2018,28(09):160.[doi:10.3969/ j. issn.1673-629X.2018.10.033]
[5]李东欣,禹 龙,田生伟,等.注意力机制的 LSTM-DBN 维语人称代词指代消解[J].计算机技术与发展,2019,29(07):33.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 007]
　LI Dong-xin,YU Long,TIAN Sheng-wei,et al.Attention Mechanism of LSTM-DBN Uyghur Personal Pronoun Anaphora Resolution[J].,2019,29(09):33.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 007]
[6]尹 鹏,周 林,郭 强,等.基于短语级注意力机制的关系抽取方法[J].计算机技术与发展,2019,29(09):24.[doi:10. 3969 / j. issn. 1673-629X. 2019. 09. 005]
　YIN Peng,ZHOU Lin,GUO Qiang,et al.Relation Extraction Based on Phrase-level Attention[J].,2019,29(09):24.[doi:10. 3969 / j. issn. 1673-629X. 2019. 09. 005]
[7]钟 诚,周浩杰,韦海亮.一种基于注意力机制的三维点云物体识别方法[J].计算机技术与发展,2020,30(04):41.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 008]
　ZHONG Cheng,ZHOU Hao-jie,WEI Hai-liang.A 3D Point Cloud Object Recognition Method Based on Attention Mechanism[J].,2020,30(09):41.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 008]
[8]王乾铭,李 吟.基于深度学习的个性化聊天机器人研究[J].计算机技术与发展,2020,30(04):79.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 015]
　WANG Qian-ming,LI Yin.Research on Personalized Chatbot Based on Deep Learning[J].,2020,30(09):79.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 015]
[9]朱立倩.基于深度学习的数显仪表字符识别[J].计算机技术与发展,2020,30(06):141.[doi:10. 3969 / j. issn. 1673-629X. 2020. 06. 027]
　ZHU Li-qian.Character Recognition of Digital Display Instrument Based on Deep Learning[J].,2020,30(09):141.[doi:10. 3969 / j. issn. 1673-629X. 2020. 06. 027]
[10]申静波,李井辉,孙丽娜.注意力机制在评论文本情感分析中的应用研究[J].计算机技术与发展,2020,30(07):169.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 036]
　SHEN Jing-bo,LI Jing-hui,SUN Li-na.Research on Application of Attention Mechanism in Comment Text Emotional Analysis[J].,2020,30(09):169.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 036]

更新日期/Last Update: 2023-09-10