Artificial neural networks to differentiate the composition and pyrolysis kinetics of fresh and long-stored maize

In this study, a novel methodology to determine plant biomass composition using artificial neural networks (ANN) is presented. This study was performed to determine the changes in the composition of fresh and 12 month-long stored biomass samples. The production of biofuels is a common method used to manage agricultural waste. However, owing to the seasonal characteristics of cultivation, storage is necessary in the production chain. The results indicated that cellulose and lignin were stable over time, with a maximum drop of 2.82 pp and 1.72 pp, respectively. Hemicellulose was determined to be less stable, with a drop of up to 9.19 pp after 12 months of storage. Regarding the kinetic parameters, the stored samples required a lower activation energy, but only for the active phase of pyrolysis. The accuracy of the proposed tool was extremely high, with a relative percentage difference as low as 12.9%.