基于时空图卷积循环神经网络的交通流预测
作者:
作者单位:

1.重庆大学自动化院;2.重庆城市管理职业学院;3.重庆大学自动化学院

作者简介:

通讯作者:

中图分类号:

TP273

基金项目:

国家自然科学基金项目(面上项目62073049);重庆市教委科学技术研究项目(KJQN202003303)


Traffic Flow Prediction Based on STG-CRNN
Author:
Affiliation:

School of Automation, Chongqing University

Fund Project:

The National Natural Science Foundation of China (General Program);Science and Technology Research Program of Chongqing Municipal Education Commission,

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对交通流预测模型中路网空间结构刻画和交通流时空特性挖掘不充分的问题,构建了一种新型的有向时空图,通过定义节点相对临近度来表征路网结构关系,通过学习邻域节点对预测节点的影响权重来表征节点间时空维度的作用关系,从而能更好表达交通流的时空特性.将时空图作为预测模型的输入,采用图卷积获取交通流数据空间依赖关系,采用门控循环神经网络获取交通流数据的时空依赖关系,建立一种基于时空图卷积循环神经网络的交通流预测模型(Spatiotemporal Graph - Convolutional Recurrent Neural Network, STG-CRNN).在美国公路交通数据集上对模型预测效果进行验证,其结果表明:STG-CRNN模型的预测结果在平均绝对误差、均方根误差和平均绝对百分误差方面,均优于自回归移动平均模型、门控循环单元模型,以及扩散卷积循环神经网络模型.

    Abstract:

    Aiming at the problem of insufficient road network spatial structure description and traffic flow spatiotemporal characteristics mining in the traffic flow prediction model, a new type of directed spatiotemporal graph is constructed, which characterizes the relationship between the road network structure by defining the relative proximity of nodes, and learning neighbor node pairs. The influence weight of the predicted node is used to characterize the relationship between the temporal and spatial dimensions of the nodes, so as to better express the temporal and spatial characteristics of traffic flow. Taking spatiotemporal graphs as the input of the prediction model, graph convolution is used to obtain the spatial dependence of traffic flow data, and the gated recurrent neural network is used to obtain the spatiotemporal dependence of traffic flow data, and a traffic flow based on the spatiotemporal graph convolution recurrent neural network is established (Spatiotemporal Graph-Convolutional Recurrent Neural Network, STG-CRNN).The model prediction effect is verified on the U.S. highway traffic data set, and the results show that the STG-CRNN model is better than the Autoregressive Moving Average Model, the Gated Recurrent Unit Model, and the Diffusion Convolutional Recurrent Neural Network Model in terms of the Mean Absolute Error, Root Mean Square Error, and Mean Absolute Percentage Error.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2020-10-20
  • 最后修改日期:2020-12-28
  • 录用日期:2021-01-11
  • 在线发布日期: 2021-02-04
  • 出版日期: