), 易清明 暨南大学 信息科学技术学院，广州 510632
收稿日期:2019-10-29出版日期:2021-02-01发布日期:2021-03-03通讯作者:蔡少委E-mail:caishaowei@stu2017.jnu.edu.cn作者简介:石敏(1984-)，女，湖北省襄阳市人，副教授，主要研究方向为信号处理、图像及图像处理研究．基金资助:国家青年科学基金(61603153);广州市产业技术重大攻关技术项目(201802010028);广州市“羊城创新创业领军人才支持计划”(2019019)A Traffic Congestion Prediction Model Based on Dilated-Dense Network
SHI Min, CAI Shaowei(), YI Qingming School of Information Science and Technology, Jinan University, Guangzhou 510632, China
Received:2019-10-29Online:2021-02-01Published:2021-03-03Contact:CAI Shaowei E-mail:caishaowei@stu2017.jnu.edu.cn摘要/Abstract
摘要： 在利用卷积神经网络模型对短时交通拥堵情况等预测场景进行预测时，由于模型的卷积池化操作过程会丢失部分数据，使得目标位置的信息出现丢失及特征的分辨率持续下降，导致模型的预测能力降低．针对此，本文提出一种空洞-稠密神经网络模型．首先，利用空洞卷积用较少的网络参数获取更大感受野的特点，充分提取出复杂多变的数据时空特征．其次，通过下采样及稠密网络的等值映射，解决参数在神经网络层数增加过程出现退化的问题．最后，取实际的城市道路平均车速数据块对网络结构的有效性进行验证．结果表明：同卷积神经网络模型相比，该网络结构预测平均绝对误差降低3%～23%.
关键词: 空洞-稠密网络, 时空特征, 卷积神经网络, 短时交通拥堵预测
Abstract: When using the convolutional neural network (CNN) model to predict short-term traffic congestion, due to the convolution pooling operation of the model, part of the data for the information of the target position will be lost, resulting in the decline of the resolution of the output features and the decrease in the predictive ability of the model. To solve this problem, this paper proposes a dilated-dense neural network model. First, it uses dilated convolution to obtain the characteristics of a larger receptive field with fewer network parameters, and fully extracts complex and variable data spatio-temporal characteristics. Then, through down-sampling and equivalent mapping of dense network, it solves the problem of parameter degradation in the process of increasing layers of neural network. Finally, it uses the actual urban road average speed data blocks to verify the validity of the model. The results show that compared with the convolutional neural network model, the average absolute error of the network structure prediction is reduced by 3% to 23%.
Key words: dilated-dense network, spatio-temporal characteristics, convolutional neural networks, short-term traffic congestion prediction
PDF全文下载地址:
点我下载PDF
