Utilização da regressão linear múltipla como ferramenta de otimização em processos industriais de moagem de trigo

Abstract

The objective of this study is to evaluate the optimization of the industrial process of wheat flour milling by using multiple linear regression. It was used data from a wheat milling industry, and the independent variables hardness, thousand kernel weight, moisture content and grain size distribution were correlated with the dependent variable extraction rate. In the study was performed a qualitative and quantitative applied research. The data used to perform multiple linear regressions were obtained in a industrial diagram with a 360 t of capacity per day, and with the following technical specifications: roller mill surface 9.0 mm/100Kg/24h, sifter area 0.050 m2 / 100 Kg / 24 h. The results inicated that the use of linear regression can assist in optimization of production process, through the management of some independent variables such as hardness, moisture content, thousand kernek weight, size distribution and their correlations with the dependent variable.