基于卷积长短时神经网络的城市轨道交通短时客流预测
作者:
作者单位:

1.西南交通大学经济与管理学院;2.江西师范大学软件学院

作者简介:

通讯作者:

中图分类号:

TP181

基金项目:

国家重点研发计划(2018YFB1601402)


Metro short-term traffic flow prediction with ConvLSTM
Author:
Affiliation:

1.School of Economics and Management,Southwest Jiaotong University;2.School of Software,Jiangxi Normal University

Fund Project:

The National Natural Science Foundation of China (General Program, Key Program, Major Research Plan)

  • 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:

    我国城市轨道交通正处在快速发展阶段,城轨交通短时客流预测对保障运营安全、优化线网结构,进而构建智慧城市具有重要意义。城轨短时客流除了具有周期性、随机性等时间特征之外,跨时段的断面客流具有相似性,并且相邻站点客流之间存在空间联系。因此本文充分考虑以上城轨短时客流的时空特征,基于卷积长短时记忆神经网络(ConvLSTM)与自适应k-means聚类算法,提出城轨短时客流预测的深度学习模型k-ConvLSTM,并设计实验对模型关键参数进行寻优。同时基于深圳市地铁IC卡的真实客流数据对模型有效性进行检验。结果表明,k-ConvLSTM在均方根误差、绝对误差均值、绝对误差百分比方面,均优于仅考虑时空特征的深度学习模型——卷积网络(CNN)与长短时记忆网络(LSTM)的并行混合模型、ConvLSTM内嵌式网络模型,以及仅考虑时间特征的深度学习模型——LSTM网络、双向长短时记忆网络(Bi-LSTM)和浅层机器学习模型——BP神经网络及支持向量回归模型(SVR)。

    Abstract:

    China’s urban rail transit is developing rapidly. Short-term passenger flow prediction is of great significance for operational safety, network optimization, and then smart city building. While the urban rail passenger flow is cyclical and random in the aspect of temporal characteristics, passenger flows in certain time slots are similar and passenger flows at adjacent stations are spatially correlated. Considering the above spatiotemporal characteristics, this research proposes a deep learning model k-ConvLSTM for urban rail short-term passenger flow prediction based on ConvLSTM and the adaptive k-means clustering algorithm. Experiments are designed to optimize the key parameters of the model. Also, in order to examine the performance of proposed model, abundant experiments are conducted based on the real passenger flow data of the Shenzhen Metro IC card. The results show that proposed k-ConvLSTM model performs better than deep learning models that only consider spatiotemporal characteristics——parallel architecture comprising convolutional network(CNN)and long short-term memory network(LSTM), ConvLSTM, and deep learning models that only consider temporal characteristics——LSTM and bi-directional long short-term memory network(Bi-LSTM), and shallow learning models——back propagation neural network(BPNN) and support vector regression model ( SVR), in terms of root mean square error, mean absolute 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:2020-04-30
  • 最后修改日期:2021-06-10
  • 录用日期:2020-08-04
  • 在线发布日期: 2020-09-02
  • 出版日期: