AN INVERSE ARTIFICIAL NEURAL NETWORK BASED MODELLING APPROACH FOR CONTROLLING HFCS ISOMERIZATION PROCESS

Abstract

***Isomerization of the glucose content of high fructose corn syrup (HFCS) into fructose needs to be strictly controlled in order to obtain a balanced product for sweetness and solubility, creating a non-trival problem. This work presents an approach to modelling of a real industrial isomerization reactor by artificial neural networks (ANN) pre-processed with principal component analysis (PCA). The initial model considered the exit fructose concentration as the output variable while the substrate flow rate to the reactor as the principal input (manipulated) variable. Then the neural network model was restructured and inversely trained by assuming the exit fructose concentration as the input variable and the feed flow rate as the output variable. Results indicate good performance by application of the developed strategy to an extensive industrial data set.