摘要
本文提出一种基于图卷积神经网络和注意力机制的短时交通流量预测模型.该模型利用图卷积网络捕获路网流量空间特征;利用自注意力机制调整网络输出,提高最终预测结果的精确度.实验结果表明,相较于对比方法,本文提出的图卷积注意力网络模型可提升预测精度.
In this paper, we propose a hybrid method with graph convolutional neural networks and attention mechanism for traffic flow prediction. The graph convolution network is used to capture the spatial features of traffic flow, and the attention mechanism is introduced to adjust and assemble the output to improve the prediction accuracy. Experimental results show that the proposed model outperforms the competing methods and can improve prediction accuracy.
作者
李志帅
吕宜生
熊刚
LI Zhishuai;LYU Yisheng;XIONG Gang(University of Chinese Academy of Sciences,Beijing 100049,China;The State Key Laboratory for Management and Control of Complex Systems,Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China)
出处
《交通工程》
2019年第4期15-19,28,共6页
Journal of Transportation Engineering
基金
中国铁路总公司科技研究开发计划课题2017X001-B
关键词
图卷积神经网络
注意力机制
短时交通流量预测
智能交通系统
graph convolutional neural network
attention mechanism
short-term traffic flow prediction
intelligent transportation systems