Accurate short-term traffic flow forecasting contributes a crucial element to the dynamic operations of traffic management and control systems. This research developed a multivariate ARIMAX model that incorporates related exogenous upstream flows {X} and an univariate SARIMA model to generate forecasts. Traffic data from the freeways A3 and A5 near Frankfurt, Germany were used for an empirical study. The estimation results show that the transfer function in the ARIMAX model gives a relationship conforming to the kinematic wave theory. The forecasting evaluations present that the multivariate ARIMAX model performs more accurate than the SARIMA model has. This result infers that the use of upstream traffic flows is effective in time series modeling for short-term traffic flow forecasting. The ARIMAX model should be used if time series analysis is adopted for short-term traffic flow forecasting.
Short-term Freeway Traffic Flow Forecasting with ARIMAX Modeling
Kurzfristige Prognose der Verkehrsstärke der Autobahn mit ARIMAX Modell
2010
Sonstige
Elektronische Ressource
Englisch
Multivariate Vehicular Traffic Flow Prediction: Evaluation of ARIMAX Modeling
Online Contents | 2001
|Multivariate Vehicular Traffic Flow Prediction: Evaluation of ARIMAX Modeling
Transportation Research Record | 2001
|