Aplicação de métodos de previsão de demanda em uma indústria de beneficiamento de cereais do Paraná

Abstract

Demand forecasting becomes an increasing necessity in a company's strategic planning seeking to remain in the global market, especially with the rising competition and competitiveness. Due the market’s fluctuation and fluctuation of consumers’ requirements and priorities, “how” and “when” to produce are questions that challenge companies every day. Wheat growth in Parana is reference all over the country and the state holds the majority of the flour mills plants. This paper presents the main quantitative methods to predict the demand of wheat production in a flour mill plant and the advantages of employing them within an organization. It also implements the Holt-Winters method to determine demand forecasting. This model deals with three variables: level, trend and seasonality of historical data. The data provided by the company, from the range of 2008 to 2013, were organized and separated into one-month periods. After data has been organized, then demand predictions, errors and smoothing constants were calculated, for both versions of Holt-Winters model: additive and multiplicative. The models were satisfactory, especially the additive one, and it can be used as a demand forecasting tool for the company.