Forecasting Travel Demand using Self-Correlated Moving Average Time Series Methods (Case Study: Qom-Tehran Freeway)

The quality of traffic flow is one of the main characteristics of the transportation network, which is widely used in issues related to urban planning, route prioritization, reducing traffic congestion and travel time; Therefore, estimating the volume of traffic and predicting it in the future is one of the important issues for transportation planners. The problem of prediction requires modeling and determining the variables affecting changes in a particular phenomenon. In this research, travel demand is predicted using time series methods. The data required for this research have been prepared from the Roads and Transportation Organization. In this study, in order to build a model, two autoregressive processes and moving average have been used. Using the above methods, the amount of demand in the coming years up to the horizon of 1404 on the Qom-Tehran freeway is predicted. The results of the study showed that among the self-correlated and moving average models and the combination of two models, namely the self-correlated moving average, the third model has a more acceptable accuracy. The parameters of this model (4,5) ARMA were obtained. Also, the validity of the constructed model, based on the average value of absolute error percentage, was 0.047, R and R2 values were calculated 0.94 and 0.89, respectively, which shows that the model has acceptable accuracy.