舰船电子工程2026，Vol.46Issue(2)：46-49,177,5.DOI:10.3969/j.issn.1672-9730.2026.02.010
基于时空注意力的ChebNet-BiGRU短时交通流预测
Short-term Traffic Flow Prediction Using ChebNet-BiGRU with Spatio-temporal Attention
摘要
Abstract
Addressing the issue that existing short-term traffic flow prediction models do not fully exploit spatio-temporal fea-tures,a ChebNet-BiGRU network model integrated with a spatio-temporal attention mechanism is proposed.This model combines the capabilities of the chebyshev network(ChebNet)for extracting spatial features and the bidirectional gated recurrent unit(BiG-RU)for extracting temporal features,while also introducing an attention mechanism to enhance the model's ability to capture key in-formation.The proposed model is tested on the PeMS04 dataset from California,USA,and a comparative analysis is conducted with other existing models.The results show that the proposed model provides a new approach to short-term traffic flow forecasting.关键词
交通流预测/ChebNet/BiGRU/注意力机制Key words
traffic flow prediction/ChebNet/BiGRU/attention mechanism分类
信息技术与安全科学引用本文复制引用
张洋明,刘利军,田保慧,任贺宇,郭虎锋..基于时空注意力的ChebNet-BiGRU短时交通流预测[J].舰船电子工程,2026,46(2):46-49,177,5.基金项目
河南交通职业技术学院2024年度校级科研项目+应用技术研究+基于时空图神经网络的短时交通流预测(编号:2024-YBXM-031) （编号:2024-YBXM-031）
河南省科技攻关项目(编号:252102241019) （编号:252102241019）
河南省高等学校重点科研项目(编号:25B510005)资助. （编号:25B510005）
