(INSTITUTO DE TECNOLOGIA, 2014) GOMES, Paulo Cesar Rocha; RODRIGUEZ, Jorge Laureano Moya; LEITE, Jandecy Cabral.; MAGALHÃES, Edilson Marques
The study evaluated different methods for forecasting the emission index of a thermoelectric power plant, using data from 2001 to 2013 and projecting values until 2020. The ARIMA, exponential, and logarithmic methods were applied, supported by tools such as Excel and SPSS20. Results showed that the ARIMA method provides more accurate forecasts for long and well-behaved time series, while linear and logarithmic methods are suitable for clear growth or decline trends. The study concluded that the choice of method depends on data characteristics and analysis objectives.
