%0 Journal Article
%T 基于ARIMA-GARCH-M模型的短时交通流预测方法<br>Short term traffic flow forecasting method based on ARIMA-GARCH-M model
%A 王晓全
%A 邵春福
%A 尹超英
%A 计寻
%A 管岭
%J 北京交通大学学报
%D 2018
%R 10.11860/j.issn.1673-0291.2018.04.011
%X 摘要 针对差分自回归移动平均(Auto-Regressive Integrated Moving Average, ARIMA)模型在获得时间序列非线性特性中的局限，基于线性递归的ARIMA模型和非线性递归的广义自回归条件异方差—均值（Generalized Autoregressive Conditional Heteroscedasticity in Mean, GARCH-M）模型，提出一种组合模型ARIMA-GARCH-M进行短时交通流预测,并利用城市快速路交通流数据进行模型预测精度的检验.结果表明：ARIMA-GARCH-M模型考虑了异方差性这一非线性特性，相比于ARIMA-SVR模型和ARIMA-GARCH模型的预测结果，本文构建模型具有较好的预测效果，能够有效提高预测精度至90.39%.<br>Abstract：Considering the drawback of ARIMA model in capturing the nonlinear characters, a hybrid ARIMA-GARCH-M model is proposed to forecast the traffic flow firstly. The proposed model combines the linear ARIMA algorithm and nonlinear GARCH-M algorithm to model the heteroscedaticity of the traffic flow time series, which can capture both the linear conditional mean and nonlinear conditional variance simultaneously. Then traffic flow data collected from the expressway in Beijing are used to verify the hybrid model. The results show that compared with the forecasting results of ARIMA-SVR model and ARIMA-GARCH model, the proposed model makes the forecasting accuracy improve significantly. The final predicting precision is 90.39%.
%K 交通工程
%K 交通流时间序列
%K 预测
%K 异方差性
%K ARIMA-GARCH-M模型&lt
%K br&gt
%K traffic engineering
%K traffic flow time series
%K forecasting
%K heteroscedaticity
%K ARIMA-GARCH-M model
%U http://jdxb.bjtu.edu.cn/CN/abstract/abstract3402.shtml