Stochastic Volatility Modeling Approach that Accounts for Uncertainties in Travel Time Reliability Forecasting

You are here: Homepage > Search

Stochastic Volatility Modeling Approach that Accounts for Uncertainties in Travel Time Reliability Forecasting

Zhang, Yanru / Haghani, Ali / Sun, Ranye

Travel time effectively measures freeway traffic conditions. Easy access to this information provides the potential to alleviate traffic congestion and to increase the reliability of road networks. However, it is a challenging task to model and estimate travel time because traffic often reveals irregular fluctuations. Traditional point prediction methods often underestimate the irregular fluctuations and provide results that may be uncertain. To capture travel time fluctuations and uncertainties associated with prediction, this paper proposes the ARIMA–stochastic volatility (ARIMA-SV) model, which generates the expected value of travel time (a point value) as well as a prediction interval. An advanced Monte Carlo Markov chain estimation method is used to fit the stochastic volatility model. Experiment results based on travel time data collected from Bluetooth detectors along an I-95 segment in Connecticut suggest that the proposed ARIMA-SV model outperforms the ARIMA–generalized autoregressive conditional heteroscedasticity model in both congested and noncongested situations. The proposed method has shown its advantages in capturing traffic fluctuations and has the potential to disseminate more reliable traffic information to travelers through advanced traveler information systems.