摘要
交通流量的准确预测对于高速路管理者进行决策至关重要;建立了小波神经网络(WNN)交通流量预测模型,并通过预测训练误差和测试误差校正预测结果来提高预测精度;首先构建WNN模型对交通流量进行初步预测,然后利用经验模态分解(EMD)和WNN模型对训练误差和测试误差进行预测;分别用训练误差预测值、测试误差预测值和两种误差预测值的加权对流量初步预测结果进行修正得到最终预测值;采用四川省成灌高速路交通流量数据进行了仿真对比实验,仿真结果表明含有误差校正的小波神经网络模型能有效提高交通流量预测精度,并且利用两种误差加权修正模型的预测精度高于利用测试误差的修正模型和利用训练误差的修正模型。
An accurate predict of traffic flow is critical for highway managers to make decisions. A wavelet neural network （WNN） model was established for forecasting traffic flow, at the same time, the prediction accuracy was improved by the training--error and test error cor rection. WNN model was established for a preliminary prediction of traffic flow, and then EMD--WNN model was proposed to forecast the train--error and test error. Finally, the correction of preliminary prediction values was carried out by predictive value of training--error, test --error and weighted value of the two kind of errors respectively. The contrastive experiments were carried out using the actual traffic flow data on Cheng--guan highway in Sichuan province. The results show that the prediction accuracy was improved by the WNN model with er- ror correction, and prediction accuracy is highest when the weighted--error-- correction is used.
出处
《计算机测量与控制》
2016年第2期168-170,181,共4页
Computer Measurement &Control
基金
四川省交通科技项目(2013c7-1)
关键词
高速路交通流量
流量预测
小波神经网络
误差预测
经验模态分解
highway traffic flow
flow prediction
wavelet neural network
error prediction
empirical mode decomposition
