Comparative Preview of Some Models for the Transportation of Grain From West of Parana Through the Rail Modal Between the Years 1997 to 2001

Abstract

Brazil has a prominent position in world grain production. With respect to the main grain like soybean, corn and wheat, the country occupies the second and third places respectively. In this context, the state of Paraná is responsible for 22% of the total of the national production, being that, 35% comes from the west region. For disposal of this production in the Paranaguá Harbour, approximately 70% of the transportation is achieved through the road modal and the remainder by railroad. Despite the advantages of the latter modal, there are many problems faced by producers who use it, among them the lack of wagons at the peak of the season. This problem may be related to the inadequate estimates on future demand for this service. Thus, this study compared the performance of three forecasting models (Holt-Winters, multiple regression and ARIMA) applied to the problem presented. The results show that, in those that the data for the grain transportation by rail modal stabilized it, the models provide a more accurate predictive capability, especially the Holt-Winters model.