Average Absolute Deviation Values Research Articles

Assessment of abiotic reduction rates of organic compounds by interpretable structural factors and experimental conditions in anoxic water environments

For organic contaminants in lake sediments, aquifers, and anaerobic bioreactors, their reduction is one of the primary transformation paths in these anoxic water environments. A simple model is introduced to predict pseudo-first order rate constants (kobs) for the abiotic reduction of organic compounds featuring diverse reducible functional groups. It utilizes the largest experimental dataset of –log kobs, encompassing 59 organic compounds (278 data points). Unlike available complex quantitative structure–activity relationship (QSAR) methods, the novel approach requires both experimental conditions and structural parameters. In comparison to one of the available general QSAR methods, the new model demonstrates favorable performance. The average absolute deviation (AAD), absolute maximum deviation (ADmax), average absolute relative deviation (AARD%), and R-squared (R2) values of the estimated outputs for 54/5 training/test data sets of the new model are 0.641/1.761, 1.761/1.417, 20.52/83.87, and 0.797/0.949, respectively. On the other hand, the available general comparative QSAR method shows the AAD: 1.311/2.301, ADmax: 3.795/3.732, AARD%: 641.0/821.2, and R2: 0.003/0.447. For the test set, AAD, AARD%, ADmax, and R2 values for the new/comparative models are 0.649/2.403, 62.20/190.5, 1.215/3.732 and 0.974/0.789, respectively. In summary, the new model offers a straightforward approach for the manual calculation of –log kobs, demonstrating excellent goodness-of-fit, reliability, precision, and accuracy.

Viscosity Modeling of Pure Alcohols Based on the Significant Structure Theory and a Cubic Equation of State

In this study, a previously developed model [Macías-Salinas, R.; Viscosity Modelling of Reservoir Fluids over Wide Temperature and Pressure Ranges. Chem. Eng. Trans. 2013, 32, 1573.] based on the significant structure theory (SST) in conjunction with the Soave–Redlich–Kwong (SRK) and Peng–Robinson (PR) cubic equations of state (CEoS) was modified for the accurate correlation of the viscosity of 25 pure alcohols from methanol to n-dodecanol, branched alcohols, diols, and triols at temperatures from 0 to 272 °C and pressures from 0.98 to 4860 bar, obtaining two final model versions (SST-SRK and SST-PR) valid for both gas and liquid phases in the case of methanol, ethanol, 2-propanol, and 2-methyl-1-propanol and only in liquid phase for the rest of alcohols considered here. In regard to n-decanol, n-undecanol, n-dodecanol, 1,3-propanediol, and 1,4-butanediol, they were studied at different temperatures but only at a pressure of 1 bar. The resulting model versions were correlated with 2418 dynamic viscosity experimental data, obtaining overall values of average absolute deviation (AAD), statistical bias, and absolute maximum deviation (AMD) of 1.4808, 0.0503, and 7.9217%, respectively, with the SST-SRK approach and 1.4799, 0.0476, and 7.4515%, respectively, with the SST-PR approach. The two modeling approaches were compared with the friction theory [Zéberg-Mikkelsen, C. K.; Viscosity Modeling of Associating Fluids Based on the Friction Theory: Pure Alcohols. Fluid Phase Equilib. 2002, 194–197, 1191.] (FT) and free volume theory [Allal, A.; A New Free Volume Model for Dynamic Viscosity and Density of Dense Fluids Versus Pressure and Temperature. Phys. Chem. Liq. 2001, 39, 1.] (FVT), using the same experimental viscosity database of this work under the same conditions of temperature and pressure, obtaining more accurate results with the SST approach for most of the modeled alcohols.

Research on personalized recommendation algorithm for micromechanical sensors based on cloud model

Abstract In order to improve the ability of micromechanical sensors to update data-matching nodes in real-time and speed up the establishment of a neighbor relationship between nodes, this paper proposes a personalized recommendation algorithm for micromechanical sensors based on the cloud model to improve the stability and real-time performance of micromechanical sensors. The algorithm uses the service attribute values of the cloud computing model and cloud clustering method to set feature term labels and establish a cloud service similarity matrix so as to meet the user-matched manufacturing service requirements. The intra-class clustering technique is applied to measure the clustering effect, and the evaluation function of each clustering number is calculated on the basis of considering the time sequence to determine the best clustering result to complete the personalized recommendation path for micromechanical sensors. To verify the application effect of the personalized recommendation algorithm for micromechanical sensors based on the cloud model, experiments are conducted. The results show that the recommendation algorithm in this paper can always control the node energy consumption below 2.5×103, and the average discovery delay is stable between 41-43 seconds. And the sensor response time is 12.1 seconds, and the average absolute deviation value is 0.23, which is nearly 1.3 times smaller than 0.53 and 0.52 of the collaborative recommendation algorithm and hybrid recommendation algorithm. It can be seen that the recommendation algorithm in this paper solves the problem of excessive neighbor discovery delay in the communication process of micromechanical sensors and effectively improves the personalized recommendation performance of micromechanical sensors.

Multi-Objective Bus Timetable Coordination Considering Travel Time Uncertainty

This paper proposes a timetable coordination method for transfer problems in a bus transit system. With a given bus network, a stochastic mixed-integer linear programming (MILP) model has been formulated to obtain coordinated bus timetables with the objective of minimizing a weighted sum of the average value of total waiting time and its average absolute deviation value, allowing for random bus travel time. The vital decision variable is the terminal departure offset time of each target bus trip within a certain off-peak period. The robust MILP model can also be used to solve the first-bus transfer problem with the introduction of several new linear constraints. A solution method based on the Monte Carlo simulation has been developed to solve the MILP model. Numerical experiments have been conducted for different scenarios. The results indicate that bus timetables coordinated by the developed model are capable of substantially reducing waiting time for transfer and non-transfer passengers. In addition, the feasibility of simplifying a common sub-route into a single transfer stop in a timetable coordination problem has been explored based on numerical experiments.

Artificial neural network and semi-empirical modeling of industrial-scale Gasoil hydrodesulfurization reactor temperature profile

The sulphur content of the desulfurization reactor product is strongly related to the severity of the reaction, which is determined by the reactor bed temperatures. In this study, the potential of neural network and semiempirical regression models to estimate temperature variation in industrial-scale gasoil HDT reactor was examined to achieve improved prediction ability. The reactor consists of 4 beds with 10 temperature indicators to control and monitor the temperature. Model developing Inlet temperature, H2/HC ratio, LCO flowrate, cracked gasoline flowrate, and straight run gasoil flowrate as input layer parameters and bed section outlet temperature as output layer was used. A database with 283 different data was built using daily records of the HDS process from an Iranian refinery. Feedforward backpropagation algorithm with learngdm learning function was used to develop ANN models. The number of neurons in the hidden layer varied between 7–13 to find the most reliable network. The optimum ANN models were selected for reactor temperature profiles on the trial-and-error method. The performance of designed models was tested by computing root mean square error (RMSE), and average absolute deviation (AAD). Comparing the neural network model and regression models, the artificial neural network model is the best model for the reactor bed temperature prediction. The value of AAD for temperatures sensors 1 to 10 in the artificial neural network model obtained 0.128, 0.094, 0.091, 0.1, 0.032, 0.067, 0.052, 0.031, 0.086, and 0.062 respectively, which show good agreement between the actual data and the artificial neural network predicted data for temperature profile in the length of the reactor.

Speed of Sound Measurements of Binary Mixtures of Hydrofluorocarbons [Pentafluoroethane (R-125), 1,1-Difluoroethane (R-152a), or 1,1,1,2,3,3,3-Heptafluoropropane (R-227ea)] with Hydrofluoroolefins [2,3,3,3-Tetrafluoropropene (R-1234yf) or trans-1,3,3,3-Tetrafluoropropene (R-1234ze(E))

Speed of sound data measured using a dual-path pulse-echo instrument are reported for three binary refrigerant mixtures, R-125/1234yf, R-1234yf/152a, and R-1234ze(E)/227ea, at compositions of (0.33/0.67) and (0.67/0.33) mole fraction. The speed of sound was studied at temperatures ranging from 230 K to 345 K from pressures slightly above the bubble point curve up to 20 MPa for the mixtures containing R-1234yf and 49 MPa for the R-1234ze(E)/227ea mixtures. The relative combined expanded speed of sound uncertainty ranged from 0.039 % to 0.317 % with a mean uncertainty over all state points of less than 0.10 %. The reported data are compared to the most recent mixture models in REFPROP for each blend studied. Comparisons of available mixture models for the R-125/1234yf and R-1234yf/152a blends exhibit average absolute deviation values ranging from 0.10 % to 0.27 %, and the average absolute deviations for R-1234ze(E)/227ea blends range from 0.62 % to 0.94 %. The comparisons show that only minor adjustments are required to the R-125/1234yf and R-1234yf/152a mixture models to represent the speed of sound data within its uncertainty. However, significant adjustments are needed to improve the current mixture models for the R-1234ze(E)/227ea blend. Deficiencies with the R-1234ze(E)/227ea mixture model are not unexpected since it (1) utilizes a pure-fluid EOS R-1234ze(E) that has been shown to inaccurately represent R-1234ze(E) speed of sound values and (2) uses binary interaction parameters for the chemically similar mixture of R-1234yf/227ea.

Development of ternary models for prediction of biogas yield in a novel modular biodigester: a case of fuzzy Mamdani model (FMM), artificial neural network (ANN), and response surface methodology (RSM)

Fuzzy Mamdani Model (FMM), Artificial Neural Network (ANN), and Response Surface Methodology (RSM) models were adopted to model and optimise biogas production on assorted substrates of poultry wastes (PW) and cow dung in a modular biodigester system. The gas produced at retention time of 42 days was within the range of 50 to 52 ml/day. The temperature range was between the range of 20 and 40 °C, while the pH values were stable at optimal range of 8.0. The statistical analysis of the experimental result obtained from methane yield, using different computational technique, reveals that fuzzy Mamdani model reflect zero (0) values for average absolute deviation (AAD), mean average error (MAE), standard error of prediction (SEP), and root-mean-square error (RMSE) and average coefficient of determination (R2) of one (1.0). ANN result obtain showed AAD value of 0.00011991, MAE value of 0.000061538, SEP value of 0.0000044369, and RMSE value of 0.00022188 and R2 value of 1.0, and RSM result reveals absolute AAD of 0.0030, MAE of 0.0015, SEP of 0.00011092, RMSE of 0.0055, and R2 of 0.9998. It can be concluded from the result of the analysis that RSM gave a good prediction of the model while ANN and FMM models produced slightly better results. The result obtained from the FMM model seems marginally better and superior in terms of forecasting of biogas and methane production rate with satisfactory outcome.

An improved simple correlation for reliable prediction of the enthalpy of fusion of cyclic and acyclic hydrocarbons including different types of saturated and unsaturated aliphatic hydrocarbons

Enthalpy (heat) of fusion of cyclic and acyclic hydrocarbons can determine molecular packing in crystals, which can be useful in correcting thermochemical data to a standard state when combined with other thermodynamic properties. An improved model is derived from a core correlation for reliable prediction of their enthalpy of fusion. The reported data are 286 for deriving and testing the new models. The calculated outputs of core and the improved correlations are compared with one of the best methods where the values of average absolute deviation (AAD) of the external set for prediction of the enthalpy of fusion by the improved correlation and comparative model are 2.24 and 3.66 kJ mol−1, respectively. Internal cross-validation of the improved correlation shows that values of leave-one-out (LOO CV) and 5-fold cross-validation (5-fold CV) are 0.972 and 0.968, respectively, which confirm that it is robust, well-behaved, and more importantly, not the over-fitted model.

A reliable model for assessment of melting points of cyclic hydrocarbons containing complex molecular structures, isomers and stereoisomers

Cyclic hydrocarbons with desirable melting points are attractive candidates as liquid fuels such as jet fuels due to the presence of high strain energy in some derivatives but available predictive methods may give large deviations for their melting points. A simple model is introduced for reliable prediction of melting points of cyclic hydrocarbons with complex molecular structures, isomers and stereoisomers that include cyclic alkane, alkene and/or alkyne, cage molecules, bridged cyclic and multicyclic hydrocarbon structures. It is based on the number of carbon atoms and two correction terms that correspond to increasing and decreasing values of the melting point resulting from specific structural parameters. Experimental data of melting points for 297 cyclic hydrocarbons were collected from different sources where data of 136 and 161 molecules are used as training and test sets, respectively. The calculated outputs of both training and test sets for the new model are compared with one of the best predictive methods. The values of average absolute deviation (AAD) of the training set for the new and comparative models are 21.0 and 52.1 K, respectively. Meanwhile, the values of AAD of the test set for these models are also 23.8 and 40.6 K, respectively. Further statistical parameters also confirm the higher reliability of the new model.

Measurement and correlation of the solubility of tin compounds and β-diketimine in selected solvents

The solubility of organotin compounds under inert atmosphere plays a significant role in homogeneous catalysis process; however, no such study has been reported. The solubilities of dimethyltin dichloride (1), stannous chloride (2), β-diketiminato ligand LH (3) (L = HC(CMeNAr)2, Ar = 2,6-Et2C6H3), and tin (II) compound [LSnCl] (4) in six organic solvents were measured using a static analytic method from 268.15 K to 313.15 K under a purified nitrogen atmosphere to select the optimum solvent for a homogeneous reaction and recrystallization. The difference in solubility can be attributed to molecular structure, coordination bonding, and molecular polarity. Six commonly used thermodynamic models were used to correlate the experimental solubility data. Some models show a great correlation with a low value of average absolute deviation (AAD). In addition, the solubility parameter of solute, partial molar excess enthalpies at infinite dilution, and thermodynamic properties were calculated to study the dissolution behavior in more depth.

Comparison of optimization approaches (response surface methodology and artificial neural network‐genetic algorithm) for a novel mixed culture approach in soybean meal fermentation

AbstractA mixed culture of Bacillus coagulans and Lactobacillus johnsonii is utilized to ferment soybean meal (SBM) and the effects of the fermentation conditions (temperature, time, and initial inoculums level) on the different antioxidative properties are measured. Both response surface methodology (RSM) and artificial neural network (ANN) are applied to the experimental data to suggest a better approach for modeling. It is observed that the fermentation conditions have significant effects on antioxidant activity, and by following a mixed culture approach, these properties can be increased efficiently. Although both ANN and RSM fitted to the experimental data with accuracy, it is evident from root mean squared error, R2, and average absolute deviation values that ANN is superior to RSM when predicting the responses. The results thus point to the advantage of using a properly trained ANN in cases of nonlinear fermentative systems and also for prediction of multiple responses simultaneously.Practical applicationsConversion of low‐cost by‐products into value‐added products has always been the main focus of fermentation industry and currently its scope is being expanded to include the use of probiotics to ferment various by‐products apart from dairy products. SBM is a promising by‐product that has not yet reached its full potential use and should be explored further to widen its range of applications. The purpose to develop optimized fermentation processes needs a suitable mathematical model that has the ability to mimic the complex nonlinearity of mixed culture fermentations without compromising its prediction capabilities. The results of this study will be further used to assess the possibility of development of a fermented‐SBM‐based supplement.

The Effects of Diethyl Carbonate in Light Naphtha Blending to Utilize New Energy Resource

Light naphtha is not reliable used as automotive fuel because of its low octane number (RON) and high vapor pressure. To utilize light naphtha as fuel, blending process may be used to obtain usable gasoline fuel. Diethyl carbonate (DEC) can be used as blending agent with light naphtha because its high octane number and low vapor pressure. Therefore in this work, the vapor pressure and octane number (RON) of DEC–light naphtha blends were measured to study the effect of DEC in light naphtha blending. The vapor pressure measurements show that vapor pressure of DEC–light naphtha blends decreased while the DEC fraction increased. The pseudo binary analysis approachment was used to correlate the vapor pressure experimental data and predict the vapor pressure of light naphtha + DEC blends. The Wilson and nonrandom two-liquid (NRTL) model are giving good correlation and prediction result with average absolute deviation (AAD) less than 1.3% meanwhile the Hildebrand & Scatchard model give quietly poor result with AAD value of 4.4%. The RON measurements show that the RON values of the blends increased while the DEC fraction increased. The DEC is successfully used as blending agent into light naphtha with 30% v/v optimum fraction.

The effect of cyano groups on the solubility of carbon dioxide in ionic liquids containing cyano groups in anion

The solubility of carbon dioxide (CO2) was investigated using three ionic liquids containing different numbers of cyano groups in the anion, 1-hexyl-3-methylimidazolium tricyanomethanide ([c6mim][C(CN)3]), 1-hexyl-3-methylimidazolium dicyanamide ([c6mim][N(CN)2]), and 1-hexyl-3-methylimidazolium thiocyanate ([c6mim][SCN]). The CO2 solubilities were determined by measuring the bubble point pressures of the CO2+ionic liquid mixtures at temperatures ranging from 303.15 to 373.15 K and pressure up to 105MPa. The experimental results showed that the solubilities of CO2 in ionic liquids increased with pressure, decreased with temperature, and increased with the length of the alkyl chains in the cation. Furthermore, a higher CO2 solubility was achieved in the ionic liquid with more cyano groups in the anion. The experimental data were correlated with the Peng-Robinson equation of state (PR-EoS), using the conventional van der Waals one fluid mixing rule and the modified Lydersen-Joback-Reid method. The average absolute deviation values of pressure (AAD-P) were 0.0658 for CO2+[c6mim][SCN], 0.0633 for CO2+[c6mim][N(CN)2], and 0.0761 for CO2+[c6mim][C(CN)3] system.

Water Quality Evaluation Using Back Propagation Artificial Neural Network Based on Self-Adaptive Particle Swarm Optimization Algorithm and Chaos Theory

To overcome the shortcomings of the traditional methods of water quality evaluation, in this paper, a novel model combines particle swarm optimization (PSO), chaos theory, self-adaptive strategy and back propagation artificial neural network (BP ANN) that was proposed to evaluate the water quality of Weihe River in China. An improved PSO algorithm with a self-adaptive inertia weight and a chaotic learning factor tuned by logistic function was developed and used to optimize the network parameters of BP ANN. The values of average absolute deviation (AAD), root mean square error of prediction (RMSEP) and squared correlation coefficient are 0.0061, 0.0163 and 0.9903, respectively. Compared with other methods, such as BP ANN, and PSO BP ANN, the proposed model displays optimal prediction performance with high precision and good correlation. The results show that the proposed method has the good prediction ability for evaluating water quality. It is convenient, reliable and high precision, which provides good analysis and evaluation method for water quality.

Comparative process optimization of pilot-scale total petroleum hydrocarbon (TPH) degradation by Paspalum scrobiculatum L. Hack using response surface methodology (RSM) and artificial neural networks (ANNs)

The aim of this study is to investigate an optimization process for the degradation of total petroleum hydrocarbon (TPH) by a tropical plant, Paspalum scrobiculatum L. Hack, using response surface methodology and artificial neural network. The optimum conditions predicted by RSM were found to be at a diesel concentration of 3%, 72 sampling days and an aeration rate of 1.77L/min with a 76.8% maximum TPH removal. The coefficients of determination (R2) and adjusted R2 for the RSM model equations were 0.8530 and 0.7208. The optimum conditions predicted by the ANN were found to be at a diesel concentration of 3%, 72 sampling days and an aeration rate of 1.02L/min with an 85.5% maximum TPH removal. Analysis using the ANN’s prediction data, which showed a higher R2 value of 0.957 and small values of Average Absolute Deviation (AAD) and Root Mean Square Error (RMSE), were 0.33% and 0.302, respectively. Validation analysis showed the predicted values by RSM and ANN were close to the validation values, whereas the ANN showed the lowest deviation, 2.57%, compared to the RSM. This finding suggests that the ANN showed a better prediction and fitting ability compared to the RSM for the non-linear regression analysis.

Avoiding binary interaction parameters in the GC-PC-SAFT model with a parametrization based in VLE and IDAC data: n-Alkanes and 1-alkanols

In this work the parameters of a PC-SAFT model combined with the group contribution of Tamouza et al. [Fluid Phase Equilib. 228–229 (2005) 409–419. http://dx.doi.org/10.1016/j.fluid.2004.10.003] were estimated using not only pure component experimental data, but also mixture vapor–liquid equilibrium (VLE) and infinite–dilution activity coefficient (IDAC) data. With the resulting parameters, the percent average absolute deviation values for vapor pressure and molar liquid volume were under 2.6% for n-alkanes and were under 7.4% for vapor pressure and 2% for molar liquid volume for 1-alkanols. For the studied mixtures, the average absolute relative deviation for bubble pressures was under 5.2%, considering 1239 experimental VLE data points in the temperature range from 273 to 422 K. For IDAC, the predictions were very similar to those obtained by other similar models. These results reveal that the proposed group contribution PC-SAFT model can be used to predict vapor–liquid equilibrium with a satisfactory accuracy for 1-alkanols and n-alkanes binary systems with no binary interaction parameter.

Ground point extraction by iterative labeling of airborne LiDAR data in a forested area

Given that the distribution of trees is irregular and individual trees differ from one another, applying a Digital Terrain Model (DTM) for cityscapes to a forest inventory causes many errors. In this study, a new DTM-generating technique that utilizes airborne Light Detection And Ranging (LiDAR), with iterative labeling for recovery of ground points, is proposed to solve the inaccuracy problem that occurs when DTMs are generated in forested areas. The proposed method consists of three steps: (1) generation of the initial DTM by a process that performs mean planar filtering and multireturn filtering, (2) recovery of ground points by iterative labeling through application of a ground extraction filter and limitation conditions, and (3) refinement to create the final DTM. The proposed method was tested at the experimental site with morphological and TerraScan DTM-generating techniques, followed by a visual assessment and a quantitative accuracy assessment through comparison with in-situ data. In the visual assessment, the proposed method exhibits such advantages as less noise and more precise representation of topographic features. Also, the method shows excellent performance in improving the average absolute deviation values of 110.3 cm and 50.4 cm over the morphological method and the TerraScan method, respectively, in the quantitative assessment. Thus, the proposed method is judged to have successfully solved the inaccuracy problem that often occurs with generation of DTMs for a forested area.

Densities, viscosities, refractive indices, and electrical conductivities of aqueous alkali salts of α-alanine

In this work, physicochemical properties such as density, viscosity, refractive index, and electrical conductivity of aqueous alkali (potassium or sodium) salts of the amino acid α-alanine (ALA), were measured at temperatures from (303.15 to 343.15)K and concentrations ranging from (0.5 to 3.5)M. Density and viscosity measurements were performed using the vibrating tube and the falling ball techniques, respectively. The refractive index at the sodium D line was measured in an automatic refractometer, while the electrical conductivity was measured using a commercial conductivity meter. An empirical equation was applied to correlate the density, refractive index, and electrical conductivity of the amino acid salt solutions with temperature and amino acid salt concentration, which gave average absolute deviation values of 0.03%, 0.01%, and 0.6%, respectively. The variation of the viscosity as a function of temperature and amino acid salt concentration was accurately represented by a modified Vogel–Tamman–Fulcher equation at an average absolute deviation of 0.5%.

Heat Capacities of Aqueous Ternary Mixtures of 2-Amino-2-Methyl-1, 3-Propanediol + Piperazine or Lithium Bromide

In this work, we reported the molar heat capacities, CP, of aqueous solutions of the sterically hindered amines 2-amino-2-methyl-1, 3-propanediol (AMPD), and its aqueous ternaries containing piperazine (PZ) or lithium bromide (LiBr). Measurements were performed in the temperature range 303.15 K to 353.15 K and atmospheric pressure by heat flow differential scanning calorimetry. The excess molar heat capacities, CP E , of the binary solutions were determined, and the data were represented as function of temperature and composition using a RedlichKister-type equation. The CP of the ternary mixtures was correlated with temperature and amine concentration using the Sohnel and Novotny equation. Results showed that the calculated data agree very well with experimental CP values at average absolute deviation values of less than 0.2%.

Prediction of solubility of CO2 in ethanol–[EMIM][Tf2N] ionic liquid mixtures using artificial neural networks based on genetic algorithm

This paper proposes artificial neural networks (ANNs) technique as a new approach to predict solubility of CO2 in ethanol–[EMIM][Tf2N] ionic liquid mixtures. The solubility equilibrium data of CO2 were measured at 0, 20, 50 and 80 and 100wt.% mixture of [EMIM][Tf2N] ionic liquid, temperatures of 313.2 and 333.2°C, and pressure range of 0–7MPa. A feed-forward multilayer perceptron (MLP) neural network with Levenberg-Marquardt learning algorithm was developed for prediction task. The ANN model was trained, validated and tested using 70%, 15% and 15% of all solubility data, respectively. An optimization procedure was performed based on genetic algorithm to select the best ANN architecture. Therefore, a three layer feed-forward neural network with Levenberg-Marquardt back-propagation training algorithm was developed and designed with four variables as inputs and one variable as output, 15 neurons in the hidden layer, log-sigmoid transfer function in the hidden layer and linear transfer function in the output layer. Based upon statistical analysis, results obtained demonstrated that there is a very little difference between predicted and experimental data of CO2 capture rate giving very low value of average absolute deviation (AAD) and high value of least square (R2) very close to one, indicating high accuracy of this model to predict output variable. The results also proved that the developed ANN model outperforms the Peng-Robinson model.