基于深度注意力模型的个体出行多步预测研究

doi:10.3969/j.issn.1674-0696.2022.10.05

重庆交通大学学报（自然科学版） ›› 2022, Vol. 41 ›› Issue (10): 35-40.DOI: 10.3969/j.issn.1674-0696.2022.10.05

• 交通+大数据人工智能 • 上一篇下一篇

基于深度注意力模型的个体出行多步预测研究

翁小雄1，任杰1，覃镇林1，罗瑞发2

(1.华南理工大学土木与交通学院，广东广州 510640； 2. 深圳市金溢科技股份有限公司，广东深圳 518000)

收稿日期:2021-03-19 修回日期:2021-08-21 发布日期:2022-10-31
作者简介:翁小雄(1958—)，女，浙江杭州人，教授，博士，主要从事大数据挖掘人类行为分析等方面的研究。E-mail：ctxxweng@qq.com 通信作者：任杰(1997—)，女，河北邯郸人，硕士，主要从事大数据挖掘人类行为分析等方面的研究。E-mail：850817872@qq.com
基金资助:
国家自然科学基金项目(51578247)

Multi-step Prediction of Individual Travel Based on Deep Attention Model

WENG Xiaoxiong1, REN Jie1, QIN Zhenlin1, LUO Ruifa2

(1. School of Civil Engineering and Transportation, South China University of Technology, Guangzhou 510640, Guangdong, China; 2. Shenzhen Jinyi Technology Co., Ltd., Shenzhen 518000, Guangdong, China)

Received:2021-03-19 Revised:2021-08-21 Published:2022-10-31

摘要/Abstract

摘要： 长期以来，对个体的出行预测一直是交通领域的研究重点。针对当前个体出行研究的局限性，提出了使用带注意力机制的序列到序列模型对个体出行进行多步预测，首先通过词嵌入的方法将个体出行特征的嵌入向量进行融合，然后基于带注意力机制的序列到序列模型设计了3种个体出行多步预测模型：整体输出式、步进输出式、多模型组合式。并将提出的模型与传统模型进行对比，最后探究了不同预测步长对实验结果带来的影响，从而验证了带注意力机制的序列到序列模型在多步预测中的适用性和优越性。

关键词: 交通运输工程；注意力机制；序列到序列模型；多步预测

Abstract: For a long time, the prediction of individual travel has been the focus of research in the field of transportation. Aiming at the limitations of current research on individual travel, a sequence-to-sequence model with attention mechanism was used to predict individual travel in multiple steps. Firstly, the embedded vectors of individual travel characteristics were fused by word embedding method. And then three kinds of individual travel multi-step prediction models were designed on the basis of the sequence to sequence model with attention mechanism, such as the overall output model, the step output model and the multi-model combined model. The proposed models were compared with the traditional model. Finally, the influence of different prediction steps on the experimental results was explored, which verified the applicability and superiority of the sequence-to-sequence model with attention mechanism in multi-step prediction.

Key words: traffic and transportation engineering; attention mechanism; sequence to sequence model; multi-step prediction

中图分类号:

U491.1+4

翁小雄1，任杰1，覃镇林1，罗瑞发2. 基于深度注意力模型的个体出行多步预测研究[J]. 重庆交通大学学报（自然科学版）, 2022, 41(10): 35-40.

WENG Xiaoxiong1, REN Jie1, QIN Zhenlin1, LUO Ruifa2. Multi-step Prediction of Individual Travel Based on Deep Attention Model[J]. Journal of Chongqing Jiaotong University(Natural Science), 2022, 41(10): 35-40.

参考文献

［1］ MO B, ZHAO Z, KOUTSOPOULOS H N, et al. Individual mobility prediction in mass transit systems using smart card data: An interpretable activity-based hidden Markov approach［J］. IEEE Transactions on Intelligent Transportation Systems, 2021, 23(8):12014-12026.
［2］ JING Y, GUO B, WANG Z, et al. Crowdtracker: Optimized urban moving object tracking using mobile crowd sensing［J］. IEEE Internet of Things Journal, 2017, 5(5): 3452-3463.
［3］ WU T, ZENG Z, QIN J, et al. An improved hmm-based approach for planning individual routes using crowd sourcing spatiotemporal data［J］. Sensors, 2020, 20(23): 6938.
［4］牟振华,李克鹏,申栋夫.基于小波降噪与贝叶斯神经网络联合模型的短时交通流量预测［J］.科学技术与工程,2020,20(33):13881-13886.
MU Zhenhua，LI Kepeng，SHEN Dongfu. Short-term traffic flow prediction based on wavelet denoising and bayesian neural network model ［J］. Science Technology and Engineering,2020,20(33):13881-13886.
［5］张玉桃. 旅客社会网络中的家庭出行预测及其行为特征分析［D］. 北京：北京交通大学, 2016.
ZHANG Yutao. Family Trip Prediction and Analysis of Behavior Char-acteristics in Passenger Social Network ［D］. Beijing: Beijing Jiaotong University, 2016.
［6］徐嘉俊. 基于GA-BP神经网络的居民出行行为预测［D］. 北京：北京交通大学, 2019.
XU Jiajun. Resident Travel Behavior Prediction Based on GA-BP Neural Network ［D］. Beijing Jiaotong University, 2019.
［7］ FERREIRA D L, NUNES B A A, CAMPOS CA V, et al. User community identification through fine-grained mobility records for smart city applications［J］. IEEE Transactions on Intelligent Transportation Systems, 2020，23(5): 4387-4401.
［8］申启东,李文勇,廉冠.基于改进马尔可夫链的交通流预测方法［J］.黑龙江交通科技,2021,44(10):178-181.
SHEN Qidong, LI Wenyong, LIAN Guan. Traffic flow prediction method based on improved markov chain communications ［J］. Science and Technology Heilongjiang, 2021,44(10):178-181.
［9］翁小雄, 郝翊. 基于LSTM引入客车占比特征的短时交通流预测［J］.重庆交通大学学报(自然科学版), 2020, 39(11): 20-25，50.
WENG Xiaoxiong, HAO Yi. Short-term traffic flow based on LSTM algorithm with the characteristics of passenger car proportion［J］. Journal of Chongqing Jiaotong University (Natural Science), 2020, 39(11): 20-25，50.
［10］肖少宁, 郭欣. 交通量预测中的神经网络模型对比分析研究［J］.物联网技术, 2015, 5(7): 39-41,45.
XIAO Shaoning, GUO Xin. Comparative analysis of neural network models in traffic volume forecasting ［J］. Internet of Things Technology, 2015, 5(7): 39-41,45.
［11］闫旭,范晓亮,郑传潘,等.基于图卷积神经网络的城市交通态势预测算法［J］.浙江大学学报(工学版),2020,54(6):1147-1155.
YAN Xu, FAN Xiaoliang, ZHENG Chuanpan, et al. Urban traffic flow prediction algorithm based on graph convolutional neural networks［J］.Journal of Zhejiang University (Engineering Science), 2020,54(6):1147-1155.

[1]	裴同松1，裴彧2. 基于马尔科夫链-BP神经网络模型对公路运量的预测研究[J]. 重庆交通大学学报（自然科学版）, 2021, 40(02): 35-41.
[2]	邬岚，王彤伟，李鹏举，翟伟. 基于窄马路理念的车道缩减设计方法[J]. 重庆交通大学学报（自然科学版）, 2020, 39(11): 26-32.
[3]	陈沿伊，曹莹，张佳楠，冯睿. 基于DEMATEL-BCC模型的深圳市生态交通发展效率评价[J]. 重庆交通大学学报（自然科学版）, 2020, 39(07): 27-32.
[4]	张兰怡，胡喜生，陈清耀，邱荣祖. 基于PCA-Logistic回归的汽车保有量预测研究[J]. 重庆交通大学学报（自然科学版）, 2017, 36(5): 104-109.
[5]	朱顺应，王红，李关寿. 路段上短时间区段内交通量预测ARIMA模型[J]. 重庆交通大学学报（自然科学版）, 2003, 22(1): 76-77.
[6]	徐建闽1，臧鹏1，首艳芳2. 基于ARIMA模型和K-means聚类分析的动态规划算法[J]. 重庆交通大学学报（自然科学版）, 2022, 41(07): 9-13.
[7]	翁小雄，覃镇林，张鹏飞. 基于图卷积门控循环网络的个体地铁出行预测[J]. 重庆交通大学学报（自然科学版）, 2022, 41(09): 1-8.

基于深度注意力模型的个体出行多步预测研究

Multi-step Prediction of Individual Travel Based on Deep Attention Model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 7

编辑推荐

Metrics