A review of some main models for traffic flow forecasting

As intelligent transportation systems (ITS) are implemented widely throughout the world, managers of transportation systems have access to large amounts of "real-time" status data. Real-time forecasting is becoming an important tool in ITS. During the past few years, various traffic-flow forecasting models have been developed and forecasting accuracy has been improved substantially. In this paper some main models for traffic flow forecasting are introduced. These models include ARIMA model, various neural network (NN) models, nonparametric model, and so on. Furthermore, in this paper some techniques for improving the forecasting power are also discussed which include judgmental adjustment technique, adaptive estimation for time-varying parameters and adopting feed-back loop structure when estimating parameters. Finally, a summary of these models and techniques is to be given.