DECISION SUPPORT SYSTEM THROUGH FUZZY LOGIC IN AN INDUSTRIAL PROCESS

The beer market is a relevant sector in Brazil. It requires producers to seek new alternatives, both in the replacement of raw materials and in the improvement of the manufacturing process. The present work aimed at applying Fuzzy Logic in the production process of Pilsen craft beer using macaúba as an adjunct. The decision support system based on fuzzy logic for the quality of the beers produced contributed to the automation of evaluation practices. Several factors influence the final quality of craft beers. Among the parameters, density, alcohol content, color, and pH stand out. Thus, the Fuzzy algorithm was developed in the Scilab® software using the Fuzzy toolbox v.4.6. The Input variables, as previously described, were: Density (Ideal and Poor), pH (Low, Ideal and High), Color (Light, Copper, and Dark) and Alcohol Content (Low, Ideal and High). The output variable of the fuzzy inference system, called Beer Quality, is composed of the answers: failed, approved with changes, and approved. The Fuzzy rule bases were described according to the knowledge of the expert in the field, in a total of 7 rules. For the beer to be approved, all the parameters under study need to be close to ideals. Through the results, when analyzing the Density and Alcohol Content, it can be observed that the quality of the beer will be considered approved when the density values are less than 1.04 and the Alcohol Content value is within the ideal range, that is, between 5 and 6%. If the Alcohol Content is outside this highlighted range, the beer will be considered approved with modification. For the Color variable, the beer will be considered approved with modification if the pH is within the ideal range (between 4 and 4.8), otherwise, the beer will be considered rejected. Thus, it can be concluded that it is possible to use fuzzy logic as a decision support system.