Application of a Single Multilayer Perceptron Model to Predict the Solubility of CO2 in Different Ionic Liquids for Gas Removal Processes

Abstract

In this work, 2099 experimental data of binary systems composed of CO2 and ionic liquids are studied to predict solubility using a multilayer perceptron. The dataset includes 33 different types of ionic liquids over a wide range of temperatures, pressures, and solubilities. The main objective of this work is to propose a procedure for the prediction of CO2 solubility in ionic liquids by establishing four stages to determine the model parameters: (1) selection of the learning algorithm, (2) optimization of the first hidden layer, (3) optimization of the second hidden layer, and (4) selection of the input combination. In this study, a bound is set on the number of model parameters: the number of model parameters must be less than the amount of predicted data. Eight different learning algorithms with (4,m,n,1)-type hidden two-layer architectures (m = 2, 4, …, 10 and n = 2, 3, …, 10) are studied, and the artificial neural network is trained with three input combinations with three combinations of thermodynamic variables such as temperature (T), pressure (P), critical temperature (Tc), critical pressure, the critical compressibility factor (Zc), and the acentric factor (ω). The results show that the 4-6-8-1 architecture with the input combination T-P-Tc-Pc and the Levenberg–Marquard learning algorithm is a very acceptable and simple model (95 parameters) with the best prediction and a maximum absolute deviation close to 10%.

Solubility of Methane in Ionic Liquids for Gas Removal Processes Using a Single Multilayer Perceptron Model

“…For example, the Peng-Robinson cubic equation of state with Kwak-Mansoori mixing rule requires three adjustable parameters for each temperature. On the other hand, works reported in the literature on gas solubilities in ILs using ANN show similar results to those obtained with conventional methods, but with the characteristic of being simple models with a reduced number of parameters [7,25].…”

“…In an MLP model, the most appropriate variables for solubility prediction have not been previously determined. However, different input combinations have been studied for solubility prediction [7,24,25,[44][45][46]. For the choice of the algorithm, a simple architecture model is used with four input variables: the experimental temperature (T) and pressure (P); and the critical temperature (Tc) and pressure (Pc).…”

“…Studies including the synthesis of new ILs, evaluation of their environmental, impact and toxicity, and technical and economic analyses, as well as experimental and theoretical research to obtain physical and physicochemical properties, appear every day in the scientific literature [16]. The phase behavior of mixtures involving ILs is also receiving increasing attention, and several studies on aspects such as liquid-liquid equilibrium, liquid-gas equilibrium, the thermodynamic consistency of data, solubility parameters, experimental difficulties, miscibility, and the effects of impurities have appeared in the literature [17][18][19][20][21][22][23][24][25].…”

“…Modeling and predicting mixture properties, including ILs, represent new and different challenges that are now possible to explore due to the availability of the necessary critical properties to apply current phase equilibrium and liquid-liquid methods. We have proposed new models to correlate and predict the solubility of gases in ILs [23][24][25][26][27][28]. The natural next step in this line of research is the study of the properties of mixtures involving ILs, such as the solubility of methane in the ILs that are analyzed in this work.…”

Optimization Analysis Model Determining PNMP Mandiri Loan Status Based on Pearson Correlation

“…The perceptron algorithm is one of the algorithms found in artificial neural networks [37]. This algorithm can provide a model that can perform better training and testing [38]. This algorithm presents a simple model in doing learning [39].…”