Blending process optimization into special fat formulation by neural networks

Computer programs are used to manage, supervise, and operate production lines of oil, margarine, butter, and mayonnaise in the fats and oils industry. Automation allows for lower-cost and better-quality products. The present paper shows a multilayer perceptron-type, second-generation neural network that was built based on a desirable product solid profile and was designed to formulate fats from three ingredients (one refined oil and two hydrogenated soybean-based storks). This network operates with three sequential decision levels, technical, availability and costs, to furnish up to nine possible formulations for the desired product, Upgrading verification was accomplished by soliciting to the formulation network all 63 products used in the upgrading (the answers were evaluated by a panel of experts and considered satisfactory) and 17 commercial products. It was possible to formulate more than 50% of the products in the network with only the three bases available. The results demonstrate the possibility of using neural networks as an alternative to the automation process for the special fats formulation process.