Short-Term Prediction of Urban Traffic Variability: Stochastic Volatility Modeling Approach

摘要: This paper addresses the problem of modeling and predicting urban traffic flow variability, which involves considerable implications for deployment dynamic transportation management systems. Traffic variability is described in terms a volatility metric, i.e., conditional variance level, as latent stochastic (low-order Markov) process. A discrete-time parametric model, referred to (SV) model employed provide short-term adaptive forecasts (speed) by using real-time detector measurements volumes occupancies an arterial. The predictive performance SV compared that generalized autoregressive heteroscedasticity (GARCH) has been recently used forecasting, with regard different measurement locations, forms data input, lengths forecasting horizon measures. results indicate potential produce out-of-sample speed significantly higher accuracy, comparison GARCH model.