Predicting the mechanical properties of glass fiber reinforced polymers via artificial neural network and adaptive neuro-fuzzy inference system

In this work, artificial neural networks (ANNs) and adaptive neuro-fuzzy inference system (ANFIS) methods were employed to make prediction on the mechanical properties of glass fiber reinforced (GFR) polymers. Therefore, toughened polyamide 6 (PA6) with various contents of maleated ethylene–propylene-rubber (EPR-g-MA) and reinforced with short glass fiber (GF) composite (PA6/EPR-g-MA/GF) were introduced to artificial neural networks (ANNs) and adaptive neuro-fuzzy inference system (ANFIS) methods. Different mechanical performances such as yield strength, Izod impact strength and modulus at different levels of feeding rates (100–200 kg/h), screw speeds (200–300 rpm) and mixing temperatures (240–260 °C) were predicted via these methods. It was shown that the obtained results through multiple inputs single output (MISO) method were very well adopted with the earlier reported experimental data including minimum errors. All the predictions of modeling results comparing with those of experimental ones had quite low root mean squared error (RMSE) values and the model performed well with R2 values.