日交通流预测的编码器-解码器深度学习模型研究

doi:10.3778/j.issn.1002-8331.2105-0022

摘要/Abstract

摘要： 精准的日交通流预测是智能交通领域的重要研究内容之一。目前已有的日交通流预测模型大多在短期预测模型的基础上通过多步预测或者多目标预测的方式改进而来。这两种改进方案中，前者对误差的传播更为敏感，而后者则忽视了预测结果的时序关系，导致预测模型精度偏低。提出了一种用于日交通流预测的编码器-解码器深度学习模型，首先将长短时记忆网络（long short-term memory，LSTM）作为编码器-解码器模型的基本单元以提高模型捕捉长期依赖关系的能力，其次引入注意力机制调节编码向量的权重以进一步提高模型的预测精度。新的模型是一种典型的序列到序列预测模型，与传统的序列到点的模型相比更加契合日交通流预测的需求。为验证模型的有效性，取美国5号州际公路西雅图段的实际交通流数据进行实验，实验结果表明，提出的预测模型在平均车流密度大于40?辆/km的时间段中，其预测结果的平均绝对百分比误差（mean absolute percentage error，MAPE）与LSTM、门控循环单元（gated recurrent unit，GRU）、反向传播（back propagation，BP）神经网络、卷积神经网络（convolutional neural network，CNN）、图卷积网络（graph convolution network，GCN）传统预测模型相比，分别减小了19%、20%、25%、16%、25%。

关键词: 日交通流预测, 编码器-解码器, 深度学习, 长短时记忆网络（LSTM）, 注意力机制

Abstract: Accurate daily traffic flow prediction is one of the important research contents in the field of intelligent transportation. At present, most of the existing daily traffic flow prediction models are improved by multi-step prediction or multi-objective prediction based on the short-term traffic flow prediction model. In these two improved schemes, the former is more sensitive to error propagation, and the latter ignores the time series relationship of prediction results, resulting in the low accuracy of the existing prediction models. This paper proposes an encoder-decoder deep learning model for daily traffic flow prediction. Firstly, long short-term memory（LSTM） neural network is taken as the basic unit of encoder-decoder model to improve the ability of the network to capture long-term dependence, and then attention mechanism is introduced to adjust the weight of coding vector to further improve the prediction accuracy of the model. Compared with the traditional sequence-to-point model, the new proposed model is a typical sequence-to-sequence model and is more suitable for the daily traffic flow prediction. In order to verify the effectiveness of the model, the actual traffic flow data on Interstate-5 in Seattle are used for the experiment. The experimental results show that the mean absolute percentage error（MAPE） of the prediction results of the proposed model is reduced by 19%, 20%, 25%, 16% and 25% respectively compared with LSTM, gated recurrent unit（GRU）, back propagation（BP） neural networks, convolutional neural network（CNN） and graph convolution network（GCN） when the average traffic density is greater than 40 veh/km.

Key words: daily traffic flow prediction, encoder-decoder, deep learning, long short-term memory（LSTM）, attention mechanism

曹阳, 茅一波, 施佺. 日交通流预测的编码器-解码器深度学习模型研究[J]. 计算机工程与应用, 2022, 58(22): 284-290.

CAO Yang, MAO Yibo, SHI Quan. Research on Encoder-Decoder Deep Learning Model for Daily Traffic Flow Prediction[J]. Computer Engineering and Applications, 2022, 58(22): 284-290.

参考文献

[1] 陆化普，孙智源，屈闻聪.大数据及其在城市智能交通系统中的应用综述[J].交通运输系统工程与信息，2015，15（5）：45-52.
LU H P，SUN Z Y，QU W C.Big data and its applications in urban intelligent transportation system[J].Journal of Transportation Systems Engineering and Information Technology，2015，15（5）：45-52.
[2] LA?A I，SER J D，VéLEZ M，et al.Road traffic forecasting：recent advances and new challenges[J].IEEE Intelligent Transportation Systems Magazine，2018，10（2）：93-109.
[3] 谭满春，李英俊，关占荣，等.周相似特性下交通流组合预测方法研究[J].计算机工程与应用，2007，43（33）：193-195.
TAN M C，LI Y J，GUAN Z R，et al.Study on hybrid approach for traffic flow prediction based on weekly similarity[J].Computer Engineering and Applications，2007，43（33）：193-195.
[4] KUMAR S V.Traffic flow prediction using Kalman filtering technique[J].Procedia Engineering，2017，187：582-587.
[5] 沈琴琴，王玥，黄悦，等.改进初值的灰色Verhulst-Markov模型及其应用[J].统计与决策，2020，36（7）：30-33.
SHEN Q Q，WANG Y，HUANG Y，et al.Grey Verhulst-Markov model with improved initial value and its application[J].Statistics & Decision，2020，36（7）：30-33.
[6] FENG X X，LING X Y，ZHENG H F，et al.Adaptive multi-kernel SVM with spatial-temporal correlation for short-term traffic flow prediction[J].IEEE Transactions on Intelligent Transportation Systems，2019，20（6）：2001-2013.
[7] 陈丹，胡明华，张洪海，等.基于贝叶斯估计的短时空域扇区交通流量预测[J].西南交通大学学报，2016，51（4）：807-814.
CHEN D，HU M H，ZHANG H H，et al.Short-term traffic flow prediction of airspace sectors based on Bayesian estimation theory[J].Journal of Southwest Jiaotong University，2016，51（4）：807-814.
[8] ZANG D，LING J W，WEI Z H，et al.Long-term traffic speed prediction based on multiscale spatio-temporal feature learning network[J].IEEE Transactions on Intelligent Transportation Systems，2019，20（10）：3700-3709.
[9] LIM D，LEE M，SEOK J.Long term traffic prediction in highway using parallel CNN[C]//IEEE 5th International Conference on Intelligent Transportation Engineering，Dalian，2020.
[10] 杨飞，方滨兴，王春露，等.基于小波和回声状态网络的交通流多步预测模型[J].吉林大学学报（工学版），2013，43（3）：646-653.
YANG F，FANG B X，WANG C L，et al.Multi-step traffic flow prediction model based on wavelet and echo state network[J].Journal of Jilin University（Engineering and Technology Edition），2013，43（3）：646-653.
[11] 袁方，舒勤，戴元.道路交通流多步预测流量控制建模研究[J].计算机仿真，2016，33（11）：152-156.
YUAN F，SHU Q，DAI Y.Study on multi-step traffic flow prediction and control model[J].Computer Simulation，2016，33（11）：152-156.
[12] MA D F，SONG X，LI P，et al.Daily traffic flow forecasting through a contextual convolutional recurrent neural network modeling inter-and intra-day traffic patterns[J].IEEE Transactions on Intelligent Transportation Systems，2021，22（5）：2627-2636.
[13] QU L C，LI W，LI W J，et al.Daily long-term traffic flow forecasting based on a deep neural network[J].Expert Systems with Applications，2019，121：304-312.
[14] 刘臣，陈静娴，郝宇辰，等.基于时空网络的地铁进出站客流量预测[J].计算机工程与应用，2021，57（18）：248-254.
LIU C，CHEN J X，HAO Y C，et al.Entrance and exit passenger flow prediction of subway stations based on spatio-temporal network[J].Computer Engineering and Applications，2021，57（18）：248-254.
[15] NGUYEN M H，NGUYEN P L，NGUYEN K，et al.PM2.5 prediction using genetic algorithm-based feature selection and encoder-decoder model[J].IEEE Acess，2021，9：57338-57350.
[16] BAHDANAU D，CHO K，BENGIO Y S.Neural machine translation by jointly learning to align and translate[J].arXiv：1409.0473，2014.
[17] 王庆荣，田可可，朱昌锋，等.融合多因素的短时交通流预测研究[J].计算机工程与应用，2021，DOI：10.3778/j.issn.1002-8331.2105-0421.
WANG Q R，TIAN K K，ZHU C F，et al.Short term traffic flow prediction based on multi-factors[J].Computer Engineering and Applications，2021，DOI：10.3778/j.issn.1002-8331.2105-0421.
[18] 桂智明，李壮壮，郭黎敏.基于ACGRU模型的短时交通流预测[J].计算机工程与应用，2020，56（21）：260-265.
GUI Z M，LI Z Z，GUO L M.Short-term traffic flow prediction based on ACGRU model[J].Computer Engineering and Applications，2020，56（21）：260-265.
[19] 李梅，宁德军，郭佳程.基于注意力机制的CNN-LSTM模型及其应用[J].计算机工程与应用，2019，55（13）：20-27.
LI M，NING D J，GUO J C.Attention mechanism-based CNN-LSTM model and its application[J].Computer Engineering and Applications，2019，55（13）：20-27.
[20] BOGAERTS T，MASEGOSA A D，ANGARITA-ZAPATA J S，et al.A graph CNN-LSTM neural network for short and long-term traffic forecasting based on trajectory data[J].Transportation Research Part C：Emerging Technologies，2020，112：62-77.