论文部分内容阅读
为了提高行程时间预测的可靠性,构建了自回归综合移动平均与广义自回归条件异方差性(ARIMAGARCH)模型进行城市主干道行程时间动态置信区间预测,其中ARIMA模型作为GARCH模型的均值方程用于捕获行程时间均值,GARCH模型用于捕获行程时间条件方差.运用昆山市交通监测系统中采集的实际交通流数据进行验证和评估.结果表明,相较于传统的ARIMA模型,提出的方法虽然不能显著提升行程时间均值的预测性能,但是在行程时间波动性预测方面具有较大的优势.该方法可捕获行程时间异方差,从而能够预测出比ARIMA模型预测的固定置信区间更能反映行程时间观测值波动性的动态置信区间.
In order to improve the reliability of travel time prediction, an autoregressive integrated moving average and generalized autoregressive conditional heteroskedasticity (ARIMAGARCH) model was constructed to predict the dynamic confidence interval of the urban arterial travel time. ARIMA model was used as the mean equation of GARCH model GARCH model is used to capture the conditional variance of travel time, and the actual traffic flow data collected in Kunshan Traffic Monitoring System are used to verify and evaluate the results. The results show that although the proposed method is not significant compared with the traditional ARIMA model Which can improve the prediction performance of travel time mean, but it has more advantages in forecasting travel time volatility.This method can capture the heteroscedasticity of travel time, and can predict the travel time observation more accurately than the fixed confidence interval predicted by ARIMA model Dynamic confidence intervals for volatility.