Premium
Expert System Based on Fuzzy Logic to Define the Production Process in the Coffee Industry
Author(s) -
HernándezVera Beatriz,
Aguilar Lasserre Alberto Alfonso,
Gastón CedilloCampos Miguel,
HerreraFranco Ligia E.,
OchoaRobles Jesús
Publication year - 2017
Publication title -
journal of food process engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.507
H-Index - 45
eISSN - 1745-4530
pISSN - 0145-8876
DOI - 10.1111/jfpe.12389
Subject(s) - expert system , production (economics) , coffee bean , fuzzy logic , quality (philosophy) , process (computing) , agricultural engineering , computer science , operations research , engineering , artificial intelligence , economics , philosophy , chemistry , epistemology , macroeconomics , operating system , organic chemistry
Coffee is the 7th largest agricultural crop in Mexico and occupies the 12th place as value generator. Cherry coffee production reaches about 1.3 million tons. This production equivalent to about 250 thousand tons of benefited coffee. Mexican exports reaches an average of 175 thousand tons, of which 90% are in grain and the rest are derivatives. The production process of coffee has not changed significantly over the years. In Mexico is common to select the type of process that coffee will be subject based on expert judgment, which causes problems in the absence of one. An expert system based on a fuzzy logic model was develop to support decision making on the type of production process considering variables such as weight, pellets, green aspect and the percentage of minor and major defects, all this information obtained from organoleptic analysis. A simulation model was used to test the validity of this expert system. Practical Applications In most Mexican coffee companies processing only a few people have the knowledge and expertise required to select the process by which green coffee or green coffee will be obtained. So an expert decides the most appropriate process to transform the coffee. The expert takes into account the customer requirements, characteristics of raw material and factors influencing the quality. But the expert cannot make decisions simultaneously when several inputs of coffee arrives and when this expert is not present the production may be delayed or even stopped. This article presents a Mandani‐Fuzzy logic model to become a support tool for selecting the most appropriate process on a dry mill according to customer requirements and in order to avoid delays in the production of coffee. The potential benefit of this tool could prevent production stoppage by the absence of an expert who can make a decision about the process to be performed and cost reduction generated by reprocessing.