基于卷积神经网络的道路交通速度预测
|
摘要
针对传统交通流预测方法由于交通流本身的非线性而使预测精度受限的问题,提出了一种基于卷积神经网络的道路交通速度预测方法。先根据道路交通速度的时间连续性和空间特征对交通数据重构出交通流的时空矩阵,作为预测模型的特征输入。然后结合卷积神经网络非线性拟合能力,及其卷积层和池化层对深层特征的抽取能力来训练模型,进而对未来的交通流做出预测。最后,使用多个指标对该方法的预测结果进行评估,表明该方法具备一定精度且能有效跟踪未来的交通流趋势。
According to the problem that accuracy of traditional traffic flow prediction methods is limited due to the nonlinear of traffic flow, a traffic speed prediction method based on conventional neural network is proposed. First, based on temporal continuity and spatial character of traffic speed of road, spatial-temporal matrix of traffic flow is constructed as the input of model. Then, combined with the nonlinear fitting ability of conventional neural network, and its conventional layers and pooling layers to extract the deep feature, the model is trained to predict the future traffic flow. Finally, evaluations using multiple indicators on prediction results of the proposed method show that it has certain accuracy and can effectively track trends of future traffic flow.
According to the problem that accuracy of traditional traffic flow prediction methods is limited due to the nonlinear of traffic flow, a traffic speed prediction method based on conventional neural network is proposed. First, based on temporal continuity and spatial character of traffic speed of road, spatial-temporal matrix of traffic flow is constructed as the input of model. Then, combined with the nonlinear fitting ability of conventional neural network, and its conventional layers and pooling layers to extract the deep feature, the model is trained to predict the future traffic flow. Finally, evaluations using multiple indicators on prediction results of the proposed method show that it has certain accuracy and can effectively track trends of future traffic flow.
引文
[1]Zhao Z,Chen W,Wu X,et al.LSTM network:a deep learning approach for short-term traffic forecast[J].IET Intelligent Transport Systems,2017,11(2):68-75.
[2]Krizhevsky A,Sutskever I,Hinton G E.Imagenet classification with deep convolutional neural networks[C]//Advances in neural information processing systems,2012:1097-1105.
[3]赵婷婷,张毅,周彧.城市路网交通流的空间互相关性[J]].清华大学学报:自然科学版,2011,51(3):313-317.
[4]LeCun Y,Bottou L,Bengio Y,et al.Gradient-based learning applied to document recognition[J].Proceedings of the IEEE,1998,86(11):2278-2324.
[5]Kingma D P,Ba J.Adam:A method for stochastic optimization[J].arXiv preprint arXiv:1412.6980,2014.
[6]Srivastava N,Hinton G,Krizhevsky A,et al.Dropout:a simple way to prevent neural networks from overfitting[J].The Journal of Machine Learning Research,2014,15(1):1929-1958.
[2]Krizhevsky A,Sutskever I,Hinton G E.Imagenet classification with deep convolutional neural networks[C]//Advances in neural information processing systems,2012:1097-1105.
[3]赵婷婷,张毅,周彧.城市路网交通流的空间互相关性[J]].清华大学学报:自然科学版,2011,51(3):313-317.
[4]LeCun Y,Bottou L,Bengio Y,et al.Gradient-based learning applied to document recognition[J].Proceedings of the IEEE,1998,86(11):2278-2324.
[5]Kingma D P,Ba J.Adam:A method for stochastic optimization[J].arXiv preprint arXiv:1412.6980,2014.
[6]Srivastava N,Hinton G,Krizhevsky A,et al.Dropout:a simple way to prevent neural networks from overfitting[J].The Journal of Machine Learning Research,2014,15(1):1929-1958.