56057

Prediction of lignin content extracted from sugarcane bagasse via hydrogen peroxide using Artificial Neural Networks (ANN)

Favorite this paper

The present work was carried out to study the application of ANN as a tool for predicting and detecting failures in a model developed to predict the percentage of lignin extracted from sugarcane bagasse after pretreatment with hydrogen peroxide (H2O2). The ANN test was performed with the results obtained by means of infrared analysis with Fourier transform (FT-IR) of the experimental samples. The experimental conditions varied from 25-45°C for the temperature and from 1.5-7.5% (v/v) for the H2O2 concentration. ANN performance was evaluated by correlation coefficient (R2) and error indexes (MSE and SSE). The values obtained for R2 and for the error indexes showed good agreement: close to 1 and close to 0, respectively.