This paper presents a time series model for daily traffic volume forecasting in rural highways. The analysis considers the characteristics of traffic variation patterns during the calendar year, i .e. seasonal and weekly periodical variations, as well as sudden changes - shocks - occurring during predictable days, e.g. Easter Holidays, Carnival.

A data set containing daily traffic volumes for the period 15/1/1985 - 31/12/1985 from a permanent count station was used to calibrate the model. The resulting model was tested in forecasting daily volumes for the time interval 1/1/1986 - 30/4/1986. Compared with the respective actual data, the results indicate that time series techniques are effective tools in developing highly accurate and inexpensive short term forecasts.

The non-linear least squares routines of statistical package RATS (Version 3.0) were used after suitable adjustments, with excellent results in relation to calculation speed and accuracy.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £78 +P&P)