CONTROL OF COFFEE (Coffea arabica L.) ESPRESSO PROCESS VARIABLES BY MULTIVARIATE STATISTICAL ANALYSIS: CONCENTRATION, TIME AND PRESSURE

Currently, Brazil is the largest exporter of coffee in the world reaching 84% of world exports between 1906 and 1907 and is a fundamental activity for the economy and development of the country. With the help of multivariate statistical analysis, a set of variables and attributes, which are responsible for coffee production, was designed to find out which ones are most relevant for the extraction of espresso. This study was based on Arabica coffee from of Minas Gerais, Brazil, with a clear medium roasting where samples were made varying the pressure, proportion and time, using multivariate statistics. The Principal Component Analysis (PCA) and Partial Least Square (PLS) analysis of the most important variables and the weights of each attribute were performed to determine the ideal values. The results of the behavior of the process variables were presented in 3D graphs. The main variables that influence the quality of espresso coffee bevarege were bitterness, astringency and sweetness. Multivariate statistics is a robust tool for analyzing variables in the coffee industry or any other industrial process that encompasses a significant number of samples and variables.