Portal de Periódicos da CAPES

Neural fuzzy model applied to autohydrolysis of Paulownia trihybrid

2010; Elsevier BV; Volume: 42; Issue: 2 Linguagem: Inglês

10.1016/j.jtice.2010.07.008

1876-1089

Minerva A.M. Zamudio, Antonio Pérez, F. López, J.C. Garcı́a, M.J. Feria, Ascensión Alfaro,

Advanced Chemical Sensor Technologies

Fuzzy modelling, which is based on the pioneering idea of Zadeh 1 , is a powerful tool for describing non-linear behaviour in complex systems.Since the 1980s, the theory of fuzzy logic has been successfully used by a number of researchers to simulate and control fermentation and anaerobic digestion processes 2 .Neural networks, which were developed by analogy with the functioning of neurons in living beings 3 , constitute one other powerful tool for modelling complex systems.The most salient feature of neural networks is their ability to exactly map non-linear behaviour via a series of input (independent variable) and output (dependent variable) data without the need for an exact knowledge of the functional relationships between the two data sets 4 .Neural networks have provided good results in the assessment of various biological systems including anaerobic digestions, stability predictions of oxidizing vegetable oils from compositions and contents in endogenous oil components, lifetime predictions in milk, predictions of trans isomer formation and changes in unsaturated fatty acids during the hydrogenation of vegetable oils, fermentation processes and kinetic analyses 5 .The combination of fuzzy logic and neural networks provides an even more useful modelling tool than either in isolation.Specifically, neural fuzzy systems have been successfully used to model organosolv delignification of raw materials such as vine shoots and tagasaste in chemical and industrial systems allowing the exploitation of lignocellulosic biomass [6][7][8] .