Predicting thermal degradation kinetics of nylon6/feather keratin blends using artificial intelligence techniques

A multi-structured architecture of artificial intelligence techniques including artificial neural network (ANN), adaptive-neuro-fuzzy-interference system (ANFIS) and radial basis function (RBF) were developed to predict thermal degradation kinetics (TDK) of nylon6 (NY6)/feather keratin (FK) blend films. By simultaneous implementation of back-propagation ANN and feed-forward ANFIS modeling on the experimental data obtained from thermogravimetric analysis (TGA) method, thermal degradation behavior of various compositions of NY6/FK blends was successfully predicted with minimum mean square errors (MSE). RBF networks were then trained on the TGA data at one heating rate for predicting analogs information at different heating rates, providing sufficient feed for TDK modeling. According to the comparison made between experimental and predicted kinetic parameters of thermal degradation process calculated from Friedman and Kissinger methods, the proposed prediction effort could effectively contribute to the estimation of precise activation energy (Ea) and reaction order (n) values with least amount of experimental work and most accuracy.