出版社:Grupo de Pesquisa Metodologias em Ensino e Aprendizagem em Ciências
摘要:The main by-products of sugarcane are sugar, anhydrous ethanol and hydrated ethanol, which are very important for the Brazilian economy. Making annual forecasts of its productions is interesting for the decision-making of the Brazilian government. Thus, the objective of this work was to use autoregressive integrated moving average (ARIMA) resources in time series, to forecast the production of sugarcane and its by-products in the Brazilian Center-South macro-region, the main producer of sugarcane, in the period from 2020 to 2025. For this purpose, historical series of the annual production of sugarcane and its by-products from 1980 to 2020 were used, in a total of four historical series, each with forty observations. In these four series, statistical tests of Pearson's correlation, of autocorrelation and partial autocorrelation and of first order differentiating were applied aiming at the good performance of forecasts using ARIMA. The forecast test was carried out in the horizon of six years later, from 2021 to 2025, for these four series. The results can be considered satisfactory, despite the mean absolute errors in percentage (MAPE) being relatively high for the four forecasts made, on average of 13.6%. As there is still no data on production of cane, sugar, anhydrous ethanol and hydrated ethanol already consolidated for the year 2021, it was not possible to make comparisons between the actual productions and the forecasts made. The importance of this work lies in providing the necessary information in agricultural planning and government policies for the agribusiness sector.
关键词:Regional development;Sugarcane production forecast;Sugar forecast;Anhydrous ethanol forecast;Hydrous ethanol production;Center-South macro-region of Brazil.