Nakata, Ichiro

Applicability of the time series analysis techniques for traffic volume forecast in developing countries - Bangkok : Asian Institute of Technology, 1986 - 51 p. - Thesis ; no. GT-85-07 . - Asian Institute of Technology. Thesis ; no. GT-85-07 .

A thesis submitted in partial fulfilment of the requirements for the degree of Master of Engineering School of Engineering and Technology

Thesis (M.Eng.) - Asian Institute of Technology, 1986

Traffic forecast is now very impormant because recent traffic increase has much influence on social problems. One of the typical forecast methods is the Box-Jenkins method. But,in developing countries it sometimes happens that the data size is insufficient for analysis. This is the result of testing the appropriacy of the Box-Jenkins method for traffic volume with small-scale data . This research establishs two kinds of method to forecast traffic volume throughout Thailand. One method uses the Box-Jenkins method directly to forecast the data in time series. The other method is a combination of the Box-Jenkins Method and a cross-sectional model. The investigation carried out through a comparison between these two methods to determine which method is more precise. The study reveals that the Box-Jenkins method is not applicable for the forecasting of small-scale traffic volume . More accurate and more detailed data collection and model improvement to decrease the forecast errors are recommended.


Traffic flow
Time-series analysis