Prediction of Degradation Temperature of Polyamide by Biofuels Using Artificial Neural Network

The aim of this study was to apply an artificial neural network to predict the degradation temperature of polyamide 12 under the influence of several factors such as environment, biofuels, and temperature over time. In this modeling, a three-layer perceptron (MLP) neural network is used. The perceptron neural network consists of three input layers, a hidden layer, and an output layer. The inputs of this network are three variables related to ethanol concentration, temperature, and time and its output is the temperature of polyamide degradation. The perceptron neural network was designed with a linear transfer function at the output layer for modeling. A comparison of experimental results and network modeling results showed R2 = 0.99. Also, using experimental data, the mean squared error and absolute percentage error of system error were 0.09 and 0.03, respectively. Then, the degradation temperature and the effects of several factors such as ethanol percent and temperatures (20, 40, and 60) degrees Celsius on the physical properties of polyamide 12 during times (0, 900, 5000, 3000, 6000, and 7000) hours are predicted and the results show the high accuracy of the neural network in estimating the degradation temperature.