深度学习交通流预测模型的现状与未来  

Current status and future prospects of deep learning models for traffic flow prediction

作　　者：张佳佳 陈静[1] ZHANG Jiajia;CHEN Jing(School of Information Technology Engineering,Tianjin University of Technology and Education,Tianjin 300222,China)

机构地区：[1]天津职业技术师范大学信息技术工程学院,天津300222

出　　处：《天津职业技术师范大学学报》2025年第1期53-58,72,共7页Journal of Tianjin University of Technology and Education

基　　金：天津市教委科研计划项目(2021KJ008);天津市津南区科技计划项目(20220105).

摘　　要：为了提高交通流预测的准确性和效率,采用文献综述方法,对深度学习模型在交通流预测领域的应用进行深入研究。研究涵盖了时空相关性构建模型的本质,包括多尺度时空特性、异质性、动态性、非线性等特点的建模过程,以及从基础深度学习模型到时空特征融合的混合神经网络等模型的发展历程。通过对比和分析,取得了对现有深度学习模型在交通预测中应用情况的全面理解,发现了在应对大规模数据、长预测范围、高计算复杂度及高维特征处理等方面的挑战和现有解决方案存在的问题。探讨了深度学习模型在交通流预测领域的发展趋势,并提出了未来发展方向,同时提供了相关开放数据集和资源以支持该领域的进一步研究。To enhance the accuracy and efficiency of traffic flow prediction based on deep learning,this paper employs a literature review approach to conduct an in-depth investigation into the application of deep learning models in traffic flow prediction.The study encompasses the essence of model construction for spatiotemporal correlation,including the modeling process of multi-scale spatiotemporal characteristics,heterogeneity,and dynamics,nonlinearity,as well as the evolution from the basic deep learning model to hybrid neural networks that integrate spatiotemporal features.Through comparative analysis,the study identifies challenges in handling large-scale data,long prediction horizons,high computational complexity,and high-dimensional feature processing,along with the limitations of existing solutions.The findings reveal the development trends of deep learning models in traffic flow prediction and propose future research directions.Additionally,relevant open datasets and resources are provided to support further advancements in this field.

关 键 词：深度学习 交通预测 交通数据 神经网络模型 

分 类 号：U491.14[交通运输工程—交通运输规划与管理]