用户名: 密码: 验证码:
基于深信度网络的城市道路网交通流预测研究
详细信息    查看官网全文
摘要
交通流预测是实现路径诱导和交通流控制的核心,应用传统神经网络模型的交通流预测方法存在容易陷入收敛到局部极小值和梯度越来越稀疏的问题。本文结合深信度网络无监督学习和有监督学习的特点,提出了一种基于深信度网络的城市道路网交通流预测研究方法。以采集后经过预处理的交通流数据为基础,用深信度网络实现道路网前5个时刻对下一时刻交通流的预测,最后对预测结果进行验证,并与传统神经网络预测结果对比分析。实验结果表明相比于其他传统神经网络方法,利用深信度网络实现交通流预测的平均绝对偏差(MAD)和平均绝对百分比误差(MAPE)更小。
Traffic flow forecasting is the core of the realization of route guidance and traffic control.The traditional neural network model to predict traffic flow is easy to converge to local minimum value and gradient is sparse.Based on the combination of the unsupervised learning and supervised learning characteristics of deep belief network,a method of research on urban traffic flow forecasting is presented.After preprocess of traffic flow data,the next time value of traffic flow is predicted according to the past 5 time data applying deep belief network.The experimental results show that this method has best performance of error of MAD and MAPE compared with methods of traditional neural network.
引文
[1]袁健,范炳全.城市道路短期交通流预测VHSSA模型[J].公路交通科技,2014,8(5):134-135.
    [2]杨兆升.新一代智能化交通控制系统关键技术及其应用[M].北京:中国铁道出版社,2008.
    [3]郑明文.径向基神经网络算法及其性能研究[D]东营:中国石油大学,2009.
    [4]姚琛,罗霞,汉克·范少伦.基于粗集和神经网络耦合的短时交通流预测[J].公路交通科技,2010,27(11):104-107.
    [5]陈文,庞琳娜.GABP神经网络在交通流预测中的应用研究[J].微计算机信息,2009,26(14):245-247.
    [6]祁伟,李晔,汪作新.季节性ARi MA模型在稀疏交通流下的预测方法[J].公路交通科技,2014,31(4):130-135.
    [7]Hinton G,Osindero S,Teh Y W.A Fast Learning Algorithm for Deep Belief Net[J].Neural Computation,2006,18(7):1527-1554.
    [8]Lee H,Grosse R,Ranganath R,et al.Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations.Proc.of the 26th Annual International Conference on Machine Learning.ACM.2009,609-616.
    [9]Hinton GE.Salakhutdinov RR.Reducing the dimensionality of data with neural networks.Science,2000,313(5786):504-507.
    [10]Dong Y,Li D.Deep learning and its applications to signal and information processing.IEEE Signal Process Mag,2001,28:145-154.
    [11]刘建伟,刘媛,罗雄麟.玻尔兹曼机研究进展[J].计算机研究与发展,2014,51(1):1-16.
    [12]Hinton G E.Training Products of Experts by Minimizing Contrastive Divergence[J].Neural Computation,2002,14(8):1771-1800.
    [13]孙志军,薛磊,许阳明.深度学习研究综述[J].计算机应用研究,2012,29(8):2806-2810.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700