A hybrid short-term traffic flow forecasting method based on spectral analysis and statistical volatility model

Abstract

Short-term traffic flow prediction is a critical aspect of Intelligent Transportation System. Timely and accurate traffic forecasting results are necessary inputs for advanced traffic management systems (ATMS) and advanced traveler information systems (ATIS). Despite the proliferation of advanced methodologies, modeling the uncertainty of traffic conditions is still a challenge, especially during congested situations. This paper presents a hybrid model for multi-step ahead traffic flow forecasting in a freeway system with real-time traffic flow data. This proposed methodology forecasts traffic flow by decomposing the data into three modeling components: an intra-day or periodic trend by introducing the spectral analysis technique, a deterministic part modeled by the ARIMA model, and the volatility estimated by the GJR-GARCH model. The aim of this study is to provide deeper insights into underlining traffic patterns and to improve the prediction accuracy and reliability by modeling these patterns separately. The forecasting performance of the proposed hybrid model is investigated with real time freeway traffic flow data from Houston, Texas. The experimental results demonstrate that the proposed method is able to unearth the underlying periodic characteristics and volatility nature of traffic flow data and show promising abilities in improving the accuracy and reliability of freeway traffic flow forecasting in multi-step ahead forecasting.