The Implementation of the Taguchi Method With Trapezoidal Fuzzy Number in the Tofu Production Process

Abstract

Indonesians consume more tofu every week, proving that it is one of the country's most well-liked and potential food ingredients. Therefore, several people benefit from this positive potential as a business opportunity and improve the quality of their products as part of a market competition strategy. This study uses the Taguchi method and fuzzy logic to optimize the multi-response characteristic tofu production process. These multi-responses include water and protein content, each of which has the characteristics of "nominal is best" and "larger is better". In this experiment, three independent variables were varied: soybean soaking time, soybean porridge boiling time, and tofu lump pressing time. The experimental design used is the orthogonal matrix L9. This study aims to determine the optimal combination of independent variables and determine the contribution of each varible to the multi-response of water content and protein content simultaneously. The findings indicated that soaking soybeans for 4 hours, boiling soybean porridge for 70 minutes, and pressing tofu lumps for 20 minutes are the ideal settings to produce optimal multi-response simultaneously. Additionally, the duration of soybeans soaking contributed 14,74%, the duration of boiling soybean porridge contributed 29,50%, and the duration of pressing lumps of tofu contributed 38,18% to the multi-response.