|本期目录/Table of Contents|

[1]徐先峰,黄刘洋,龚 美.基于卷积神经网络与双向长短时记忆网络组合模型的短时交通流预测[J].工业仪表与自动化装置,2020,(01):13-18.
　XU Xianfeng,HUANG Liuyang,GONG Mei.Short-term traffic flow prediction based on combined model of convolutional neural network and bidirectional long-term memory network[J].Industrial Instrumentation & Automation,2020,(01):13-18.
点击复制

基于卷积神经网络与双向长短时记忆网络组合模型的短时交通流预测

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2020年01期
页码:
13-18
栏目:
出版日期:
2020-02-15

文章信息/Info

Title:
Short-term traffic flow prediction based on combined model of convolutional neural network and bidirectional long-term memory network
作者:
徐先峰; 黄刘洋; 龚 美
长安大学 电子与控制工程学院，西安 710064
Author(s):
XU Xianfeng;  HUANG Liuyang;  GONG Mei
School of Electronic and Control Engineering, Chang’an University, Xi’an 710064, China
关键词:
智能交通; 短时交通流预测; 深度学习; CNN; BiLSTM
Keywords:
intelligent transportation;  short-term traffic flow prediction;  deep learning;  CNN;  BiLSTM
分类号:
U491.1+4
DOI:
-
文献标志码:
A
摘要:
及时准确的交通流信息对于智能交通系统的实现至关重要。针对现有预测方法不能充分利用交通流的时空特征,进而不能很好地提取交通流序列内在规律的问题,提出了一种结合卷积神经网络(CNN)和双向长短时记忆网络(BiLSTM)的深度学习预测模型(C-BiLSTM),在网络底层利用一维CNN来捕获观测点交通流数据的空间特征,然后输入到双向LSTM网络提取时间周期特征,最后由全连接层输出预测结果。使用美国交通研究数据实验室的实测交通数据进行验证,结果表明,所提出的C-BiLSTM组合模型具有更高的准确性,其性能相比双向LSTM网络预测模型提升了1.6%,相比单向LSTM网络预测模型提升了6.6%,是一种高精度的交通流预测模型。
Abstract:
Timely and accurate traffic flow information is critical to the realization of intelligent transportation systems.In view of the fact that the existing prediction methods can not fully utilize the temporal and spatial characteristics of traffic flow,and can not extract the inherent law of traffic flow sequence well,a combination of convolutional neural network(CNN) and bidirectional long-term memory network(BiLSTM) is proposed.The deep learning prediction model(C-BiLSTM) has the advantage of using the one-dimensional CNN to capture the spatial characteristics of the observation point traffic flow data at the bottom of the network.The advantages are then input to the bidirectional LSTM network to extract the time period features,and finally the full connection layer outputs the prediction results.Validated traffic data from the US Transportation Research Data Laboratory was used for verification.The results show that the proposed C-BiLSTM combination model has higher accuracy, and its performance is improved by 1.6% compared with the bidirectional LSTM network prediction model.Compared with the LSTM network prediction model, it is increased by 6.6%.It is a high- precision traffic flow prediction model.

参考文献/References:

[1] Kumar K, Parida M. Short term traffic flow prediction in heterogeneous condition using artificial neural network[J]. Transport, 2015,30(4): 397-405.

[2] Yang H,Dillon T,Chen Y. Optimized Structure of the Traffic Flow Forecasting Model with a Deep Learning Approach[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 28(10): 2371–2381.
[3] Voort M,Dougherty M,Watson S. Combining kohonen maps with arima time series models to forecast traffic flow[J]. Transportation Research Part C Emerging Technologies, 1996, 4(5): 307-318.
[4] Edward Y, Zygmunt J. An Algorithm for Prediction of Link Lifetime in MA-NET Based on Unscented Kalm Filter[J]. IEEE Communications Letters, 2009,13(10): 782-784.
[5] Huang W,Song G,Hong H,et al.Deep Architecture for Traffic Flow Prediction: Deep Belief Networks With Multitask Learning[J].IEEE Transactions on Intelligent Transportation Systems, 2014, 15(5): 2191-2201.
[6] 多丽,刘义艳,程绍峰.基于逆传播算法的十字路口交通流预测[J].工业仪表与自动化装置,2017(05): 120-123.
[7] K Chan,T Dillon,J Singh, et al. Neural-Network-Based Models for Short-Term Traffic Flow Forecasting Using a Hybrid Exponential Smoothing and Levenberg-Marquardt Algorithm[J].IEEE Transactions on Intelligent Transportation Systems, 2012(13): 644-654.
[8] Jin Y, Tan E. Hybrid Traffic Forecasting Model With Fusion of Multiple Spatial Toll Collection Data and Remote Micro- wave Sensor Data[J].IEEE Access, 2018(6):79211- 79221.
[9] 罗文慧,董宝田,王泽胜.基于CNN-SVR混合深度学习模型的短时交通流预测[J].交通运输系统工程与信息,2017, (17)5: 69-74.
[10] 王青松,谢兴生,佘颢.基于CNN-XGBoost混合模型的短时交通流预测[J].测控技术,2019,38(4): 37-41.
[11] Chiou H,Ping K.A Deep CNN-LSTM Model for Particulate Matter(PM2.5) Forecasting in Smart Cities[J]. Sensors, 2018(18): 2220-2242.
[12] Zhang D, Kabuka M. Combining weather condition data to predict traffic flow:A GRU-based deep learning approach [J].IET Intelligent Transport Systems,2018,12(7): 578-585.
[13] Lü Y, Duan Y, Kang W, et al. Traffic Flow Prediction With Big Data:A Deep Learning Approach[J].IEEE Transactions on Intelligent Transportation Systems,2015,16(2):865-873.

相似文献/References:

[1]冯宇平,安雪美,李树光.基于情景感知的智能交通CPS研究[J].工业仪表与自动化装置,2017,(05):124.
　FENG Yuping,AN Xuemei,LI Shuguang.Research on intelligent transportation CPS based on context-aware[J].Industrial Instrumentation & Automation,2017,(01):124.

备注/Memo

备注/Memo:
收稿日期：2019-06-18
基金项目：国家自然科学基金(61201407,61473047);陕西省自然科学基础研究计划(2016JQ5103,2019GY-002);长安大学中央高校基本科研业务费(300102328202);西安市智慧高速公路信息融合与控制重点实验室项目(ZD13CG46)
作者简介：徐先峰(1982)，男，山东宁阳人，博士，副教授，硕士生导师，主要研究方向为盲信号处理及其应用，智能交通。
更新日期/Last Update: 2020-01-10