Estudo de métodos de previsão de demanda aplicados a cavacos de madeira para um digestor de celulose

Abstract

This project was carried out with given data from a pulp mill, where after an initial analysis of historical demand, it was verified that the stock pile of wood chips before the digester could be oversized. The objective of this project was the development of a model capable of forecasting the demand for the equipment, thus reducing the quantity of raw material. From this, it is also expected to maintain the quality of the chips, which will be stored for a shorter time, and in addition to ensuring uninterrupted and irreducible supply during maintenance periods in the production line prior to the bottleneck equipment. For this, the historical demand data of the equipment and the information on maintenance of the previous ones were analyzed, allowing the definition of the best forecasting model for the case. Finally, it was achieved the best forecasting model for the current demand, evaluated by its efficiency, which will allow the company to take better decisions about the maintenance schedule.