Short-Term Prediction of Traffic Volume in Urban Arterials

Abstract

This paper attempts to develop time-series models for forecasting traffic volume in urban arterials. The Box-Jenkins approach is used to estimate the time-series models. A 1-min data set representing traffic volume on five major urban arterials were available to estimate time-series models. The Box-Jenkins autoregressive integrated moving average (ARIMA) model of order (0, 1, 1) turned out to be the most adequate model in reproducing all original time series. The developed model is easy to understand and implement. Further, the model is computationally tractable, and only requires the storage of the last forecasted error and current traffic observation.