Evaluation of Methane Adsorption by Mineral Adsorbents Using Neural Network Coding in MATLAB

Abstract

This research analyzed methane adsorption data on inorganic adsorbents using neural network modeling. In this research, various data obtained by different researchers for the selective methane absorption by the neural network in MATLAB software were analyzed. The optimal points were identified with the help of a correct algorithm, layers and precise optimization functions. The investigated method in the neural network uses a perceptron network to model the methane gas separation process. The laboratory data of experimental results of adsorption methane has been used and entered as the input experimental data to the MLP multi-layer neural network modeling. According to the created results and the evaluation of the regression coefficient R2 and MSE values, the Levenberg-Marquart training function has been selected as the optimal training function. The tansig transfer function with R2=0.98995 and the number of neurons 30 was chosen as the optimal transfer function.