辽宁工程技术大学学报（自然科学版）2023，Vol.42Issue(6)：763-768,6.DOI:10.11956/j.issn.1008-0562.2023.06.017
结合时间注意力机制的Bi-GRU-Atten的短时交通流预测
Short-term traffic flow forecast based on Bi-GRU-Atten algorithm with multi-layer time attention mechanism
摘要
Abstract
In order to use the deep learning model to predict the future highway traffic flow,the bi-directional gated loop unit algorithm(Bi-GRU)is used to extract information from two-way propagation to fully learn the time correlation characteristics of historical traffic flow,at the same time,the attention mechanism is adopted to distinguish the importance of traffic time series by correctly allocating weights,so as to further improve the computational efficiency of prediction.The open source highway data set is used to verify the model,and the results show that the proposed method is superior to other deep learning algorithms in computational efficiency and prediction accuracy,such as recurrent neural network(RNN),long-term and short-term memory network(LSTM),bidirectional long-and short-term memory network(Bi-LSTM),and bidirectional gated cycle unit algorithm without attention mechanism,and can be used to predict short-term traffic flow.关键词
交通流量预测/时间注意力机制/Bi-GRU/时间相关特征/预测效率Key words
traffic flow forecast/time attention mechanism/Bi-GRU/time-dependent feature/prediction efficiency分类
信息技术与安全科学引用本文复制引用
徐厚生,郭佳丽..结合时间注意力机制的Bi-GRU-Atten的短时交通流预测[J].辽宁工程技术大学学报（自然科学版）,2023,42(6):763-768,6.基金项目
国家自然科学基金项目(61803275) （61803275）
辽宁省"兴辽英才计划"项目(XLYC1907044) （XLYC1907044）
辽宁省自然科学基金项目(2020-MS-218) （2020-MS-218）
辽宁省教育厅重点攻关项目(LJKZZ20220082) （LJKZZ20220082）
