Multilayer perceptron prediction for the production of methyl acetate via reactive distillation

Abstract EN

In this paper, it was given a profile for the design of reactive distillation to produce methyl acetate from acetic acid and methanol in the presence of sulfuric acid as catalyst.

The conversion of the reaction was studied as function to time, reflux ratio and feed mole ratio and predicted using Multilayer Perceptron through SPSS software 21.

The result shows that using ANN will reduce the Squares Error to 0.262 and the Relative Error to 12.1%.