Modeling and Optimization of Anode‐Supported Solid Oxide Fuel Cells on Cell Parameters via Artificial Neural Network and Genetic Algorithm

An artificial neural network (ANN) and a genetic algorithm (GA) are employed to model and optimize cell parameters to improve the performance of singular, intermediate‐temperature, solid oxide fuel cells (IT‐SOFCs). The ANN model uses a feed‐forward neural network with an error back‐propagation algorithm. The ANN is trained using experimental data as a black‐box without using physical models. The developed model is able to predict the performance of the SOFC. An optimization algorithm is utilized to select the optimal SOFC parameters. The optimal values of four cell parameters (anode support thickness, anode support porosity, electrolyte thickness, and functional layer cathode thickness) are determined by using the GA under different conditions. The results show that these optimum cell parameters deliver the highest maximum power density under different constraints on the anode support thickness, porosity, and electrolyte thickness.     

Artificial Neural Network Modeling of Hydrodynamics of Liquid‐Solid Circulating Fluidized Beds

Solids holdup and solids circulation rate are the two important hydrodynamic variables affected by process conditions. These two variables have a significant influence on the performance of a liquid‐solid circulating fluidized bed (LSCFB). An artificial neural network (ANN) methodology was developed and simulated to predict the performance of the LSCFB for the experimental dataset collected under various process conditions. Different statistical parameters were applied to evaluate the prominent and unique characteristic features of the ANN‐predicted parameters. The ANN model successfully predicted the experimental observations and captured the actual nonlinear behavior noticed during the experiments. Model validation confirmed that this data‐driven technique can be used to model such nonlinear systems.     

UNIFICATION OF FRUIT WATER SORPTION ISOTHERMS USING ARTIFICIAL NEURAL NETWORKS

The chemical composition, water activity, temperature and equilibrium moisture content (EMC) for 10 selected fruits were determined. Two methods of water sorption modeling, the GAB equation and the artificial neural network (ANN) method, were compared for their ability to predict water sorption behavior. Unlike the GAB equation, which uses only physical data for modeling, the ANN method uses both physical and chemical compositional data to make predictions. The ANN was superior, in most cases, to that of the GAB equation, in predicting EMC. This superiority was due to the availability of the additional chemical compositional information. The ANN method could predict EMC with a mean relative error of 9.85% and a standard error (S_x) of 1.59% EMC. The correlation coefficient (r²) of the relationship between the actual and predicted values of equilibrium moisture content obtained by the ANN was 0.9938. The ANN model was able to show a temperature dependent crossing of water sorption isotherms, due to the dissolution of sugar crystals in the fruit. The ANN was also able to predict the extent of crossing, depending upon differences in the individual fruit chemical composition.     

Artificial neural networks model for the prediction of steady state phenol biodegradation in a pulsed plate bioreactor

BACKGROUND: A recent innovation in fixed film bioreactors is the pulsed plate bioreactor (PPBR) with immobilized cells. The successful development of a theoretical model for this reactor relies on the knowledge of several parameters, which may vary with the process conditions. It may also be a time‐consuming and costly task because of their nonlinear nature. Artificial neural networks (ANN) offer the potential of a generic approach to the modeling of nonlinear systems. RESULTS: A feedforward ANN based model for the prediction of steady state percentage degradation of phenol in a PPBR by immobilized cells of Nocardia hydrocarbonoxydans (NCIM 2386) during continuous biodegradation has been developed to correlate the steady state percentage degradation with the flow rate, influent phenol concentration and vibrational velocity (amplitude × frequency). The model used two hidden layers and 53 parameters (weights and biases). The network model was then compared with a Multiple Regression Analysis (MRA) model, derived from the same training data. Further these two models were used to predict the percentage degradation of phenol for blind test data. CONCLUSIONS: The performance of the ANN model was superior to that of the MRA model and was found to be an efficient data‐driven tool to predict the performance of a PPBR for phenol biodegradation. Copyright © 2008 Society of Chemical Industry     

UNIFICATION OF FRUIT WATER SORPTION ISOTHERMS USING ARTIFICIAL NEURAL NETWORKS

The chemical composition, water activity, temperature and equilibrium moisture content (EMC) for 10 selected fruits were determined. Two methods of water sorption modeling, the GAB equation and the artificial neural network (ANN) method, were compared for their ability to predict water sorption behavior. Unlike the GAB equation, which uses only physical data for modeling, the ANN method uses both physical and chemical compositional data to make predictions. The ANN was superior, in most cases, to that of the GAB equation, in predicting EMC. This superiority was due to the availability of the additional chemical compositional information. The ANN method could predict EMC with a mean relative error of 9.85% and a standard error (S _x ) of 1.59% EMC. The correlation coefficient (r ²) of the relationship between the actual and predicted values of equilibrium moisture content obtained by the ANN was 0.9938. The ANN model was able to show a temperature dependent crossing of water sorption isotherms, due to the dissolution of sugar crystals in the fruit. The ANN was also able to predict the extent of crossing, depending upon differences in the individual fruit chemical composition.     

Performance evaluation of an ANN–GA aided experimental modeling and optimization procedure for enhanced synthesis of marine biosurfactant in a stirred tank reactor

BACKGROUND: An improved resilient back‐propagation neural network modeling coupled with genetic algorithm aided optimization technique was employed for optimizing the process variables to maximize lipopeptide biosurfactant production by marine Bacillus circulans. RESULTS: An artificial neural network (ANN) was used to develop a non‐linear model based on a 2⁴ full factorial central composite design involving four independent parameters, agitation, aeration, temperature and pH with biosurfactant concentration as the process output. The polynomial model was optimized to maximize lipopeptide biosurfactants concentration using a genetic algorithm (GA). The ranges and levels of these critical process parameters were determined through single‐factor‐at‐a‐time experimental strategy. Improved ANN‐GA modeling and optimization were performed using MATLAB v.7.6 and the experimental design was obtained using Design Expert v.7.0. The ANN model was developed using the advanced neural network architecture called resilient back‐propagation algorithm. CONCLUSION: Process optimization for maximum production of marine microbial surfactant involving ANN‐GA aided experimental modeling and optimization was successfully carried out as the predicted optimal conditions were well validated by performing actual fermentation experiments. Approximately 52% enhancement in biosurfactant concentration was achieved using the above‐mentioned optimization strategy. © 2012 Society of Chemical Industry     

Designing and Testing a Chemical Demulsifier Dosage Controller in a Crude Oil Desalting Plant: An Artificial Intelligence‐Based Network Approach

The aim of this paper is to present an artificial neural network (ANN) controller trained on a historical data set that covers a wide operating range of the fundamental parameters that affect the demulsifier dosage in a crude oil desalting process. The designed controller was tested and implemented on‐line in a gas‐oil separation plant. The results indicate that the current control strategy overinjects chemical demulsifier into the desalting process whereas the proposed ANN controller predicts a lower demulsifier dosage while keeping the salt content within its specification targets. Since an on‐line salt analyzer is not available in the desalting plant, an ANN based on historical measurements of the salt content in the desalting process was also developed. The results show that the predictions made by this ANN controller can be used as an on‐line strategy to predict and control the salt concentration in the treated oil.     

Optimization of acrylic dry spinning production line by using artificial neural network and genetic algorithm

Acrylic fibers are synthetic fibers with wide applications. A couple of methods can be utilized in their manufacture, one of which is the dry spinning process. The parameters in this method have nonlinear relationships, making the process very complex. To the best of the authors' knowledge, no comprehensive study has yet been conducted on the optimization of acrylic dry spinning production using computer algorithms. In this study, such parameters as extruder temperature in and around the head, solution viscosity, water content in the solution, formic acid content of the solution, and the retention time of the solution in the reactor were measured in an attempt to predict the behavior of the dry spinning process. The color index of the manufactured fibers was used as an indicator of production quality and statistical methods were employed to determine the parameters affecting the process. An artificial neural network (ANN) using the back propagation training algorithm was then designed to predict the color index. ANN parameters including the number of hidden layers, number of neurons in each layer, adaptive learning rate, activation functions, number of max fail epochs, validation and test data were optimized using a genetic algorithm (GA). The trial and error method was used to optimize the GA parameters like population size, number of generations, crossover or mutation rates, and various selection functions. Finally, an ANN with a high accuracy was designed to predict the behavior of the dry spinning process. This method is capable of preventing the manufacturing of undesired fibers. © 2010 Wiley Periodicals, Inc. J Appl Polym Sci, 2011     

A developed smart technique to predict minimum miscible pressure—eor implications

Miscible gas injection (MGI) processes such as miscible CO₂ flooding have been in use as attractive EOR options, especially in conventional oil reserves. Optimal design of MGI is strongly dependent on parameters such as gas–oil minimum miscibility pressure (MMP), which is normally determined through expensive and time‐consuming laboratory tests. Thus, developing a fast and reliable technique to predict gas–oil MMP is inevitable. To address this issue, a smart model is developed in this paper to forecast gas–oil MMP on the basis of a feed‐forward artificial neural network (FF‐ANN) combined with particle swarm optimisation (PSO). The MMP of a reservoir fluid was considered as a function of reservoir temperature and the compositions of oil and injected gas in the proposed model. Results of this study indicate that reservoir temperature among the input parameters selected for the PSO–ANN has the greatest impact on MMP value. The developed PSO–ANN model was examined using experimental data, and a reasonable match was attained showing a good potential for the proposed predictive tools in estimation of gas–oil MMP. Compared with other available methods, the proposed model is capable of forecasting oil–gas MMP more accurately in wide ranges of thermodynamic and process conditions. All predictive models used other than the PSO–ANN model failed in providing a good estimate of the oil–gas MMP of the hydrocarbon mixtures in Azadegan oilfield, Iran. © 2013 Canadian Society for Chemical Engineering     

Prediction of Gas‐Phase Adsorption Isotherms Using Neural Nets

This study investigated a number of models (the modified Sips', Dubinin‐Astakhov's, VSM theory, the generalized Khan et al.'s model and a simple artificial neural network (ANN)) to predict the effect of temperature on equilibrium adsorption of hydrocarbon gases and vapors on activated carbon. Published data on the adsorption of methane, ethane and propane on activated carbon at 311 K to 505 K were used to estimate the parameters of the conventional models and train the network. Then, the conventional models and the ANN were used to predict the isotherm at a single temperature for each adsorbate, and these results were compared with experimental data. It was found that the ANN model had a lower mean relative error than the conventional models.     

Prediction of gas‐phase adsorption isotherms using neural nets

This study investigated a number of models (the modified Sips', Dubinin‐Astakhov's, VSM theory, the generalized Khan et al.'s model and a simple artificial neural network (ANN)) to predict the effect of temperature on equilibrium adsorption of hydrocarbon gases and vapors on activated carbon. Published data on the adsorption of methane, ethane and propane on activated carbon at 311 K to 505 K were used to estimate the parameters of the conventional models and train the network. Then, the conventional models and the ANN were used to predict the isotherm at a single temperature for each adsorbate, and these results were compared with experimental data. It was found that the ANN model had a lower mean relative error than the conventional models.     

Experimental study and artificial neural network simulation of methane adsorption on activated carbon

The adsorption of methane on two activated carbons with different physical properties was measured. Adsorption isotherms were obtained by static volumetric method at different temperatures and pressures. The experimental results sow the best gas storage capacity was 113.5 V/V at temperature 280 K and pressure 8.5MPa on an activated carbon with surface area 1,060 m²/gr. An artificial neural network (ANN) based on genetic algorithm (GA) was used to predict amount of adsorption. The experimental data including input pressure, temperature and surface area of adsorbents as input parameters were used to create a GA-ANN simulation. The simulation results were compared with the experimental data and a good agreement was observed. The simulation was applied to calculate isosteric heat of adsorption by using the Clausius-Clapeyron equation. Comparison of the calculated adsorption heat showed different surface heterogeneity of the adsorbents.     

Estimation of breakthrough time for water coning in fractured systems: Experimental study and connectionist modeling

Water coning in petroleum reservoirs leads to lower well productivity and higher operational costs. Adequate knowledge of coning phenomena and breakthrough time is essential to overcome this issue. A series of experiments using fractured porous media models were conducted to investigate the effects of production process and pore structure characteristics on water coning. In addition, a hybrid artificial neural network (ANN) with particle swarm optimization (PSO) algorithm was applied to predict breakthrough time of water coning as a function of production rate and physical model properties. Data from the literature combined with experimental data generated in this study were used to develop and verify the ANN‐PSO model. A good correlation was found between the predicted and real data sets having an absolute maximum error percentage less than 9%. The developed ANN‐PSO model is able to estimate breakthrough time and critical production rate with higher accuracy compared to the conventional or back propagation (BP) ANN (ANN‐BP) and common correlations. The presence of vertical fractures was found to accelerate considerably the water coning phenomena during oil production. Results of this study using combined data suggest the potential application of ANN‐PSO in predicting the water breakthrough time and critical production rate that are critical in designing and evaluating production strategies for naturally fractured reservoirs. © 2014 American Institute of Chemical Engineers AIChE J, 60: 1905–1919, 2014     

Comparison between the artificial neural network,SAFT and PRSV approach in obtaining the solubility of solid aromatic compounds in supercritical carbon dioxide

The aim of this study is to model the solubilities of solid aromatic compounds in supercritical carbon dioxide (SCCO₂) using feed-forward artificial neural network (ANN). Temperature, pressure, critical properties and acentric factor of each solute have been used as independent variables of ANN model. The parameters of multi-layer perceptron (MLP) network have been adjusted by back propagation learning algorithm using experimental data which have been collected from various literatures. In order to find the optimal topology of the MLP, different networks were trained and examined and the network with minimum absolute average relative deviation percent (AARD%), mean square error (MSE) and suitable regression coefficient (R²) has been selected as an optimal configuration. By this procedure a single hidden layer network composed of nineteen hidden neurons has been found as an optimal topology. Sensitivity error analyses confirmed that the optimal ANN can predict experimental data with an excellent agreement (AARD% = 4.99, MSE = 7.08 × 10⁻⁷ and R² = 0.99699). Capability of the proposed ANN model has compared with those published results which have obtained by SAFT combined with eight different mixing rules (one, two and three parameters mixing rules) and PRSV equation of state (EOS). The best presented overall AARD% for SAFT approach with one, two and three parameters mixing rules are 16.15, 12.32% and 7.65%, respectively while PRSV EOS showed AARD% of 21.10%. The results emphasize that the proposed ANN model can predict the solubilities of solid aromatic compounds in SCCO₂ more accurate than SAFT and PRSV EOS.     

Using artificial neural networks to model and interpret electrospun polysaccharide (Hylon VII starch) nanofiber diameter

Present work was aimed to develop an artificial neural networks (ANN) model to predict the polysaccharide-based biopolymer (Hylon VII starch) nanofiber diameter and classification of its quality (good, fair, and poor) as a function of polymer concentration, spinning distance, feed rate, and applied voltage during the electrospinning process. The relationship between diameter and its quality with process parameters is complex and nonlinear. The backpropagation algorithm was used to train the ANN model and achieved the classification accuracy, precision, and recall of 93.9%, 95.2%, and 95.2%, respectively. The average errors of the predicted fiber diameter for training and unseen testing data were found to be 0.05% and 2.6%, respectively. A stand-alone ANN software was designed to extract information on the electrospinning system from a small experimental database. It was successful in establishing the relationship between electrospinning process parameters and fiber quality and diameter. The yield of smaller diameter with good quality was favored by lower feed rate, lower polymer solution concentration, and higher applied voltage.     

人工神经网络用于 PBX 炸药装药密度的研究

比较了利用人工神经网络（ＡＮＮ）和混合物密度计算公式对一种聚合物粘结炸药（ＰＢＸ）的装药密度计算的结果。发现利用人工神经网络所得的结果可以正确反映配方组成和温度与装药密度的关系，而且对于新配方和温度，可以利用人工神经网络得到正确的密度数值预报。     

A sequence optimization strategy for chromatographic separation in reversed‐phase high‐performance liquid chromatography

A sequence optimization strategy combining an artificial neural network (ANN) and a chromatographic response function (CRF) for chromatographic separation in reversed‐phase high‐performance liquid chromatography has been proposed. Experiments were appropriately designed to obtain unbiased data concerning the effects of varying the mobile phase composition, flow‐rate, and temperature. The ANN was then used to simultaneously predict the resolution and analysis time, which are the two most important features of chromatographic separation. Subsequently, a CRF consisting of resolution and analysis time was used to predict the optimum operating conditions for different specialized purposes. The experimental chromatograms were consistent with those predicted for given conditions, which verified the applicability of the method. Furthermore, the proposed optimization strategy was applied to literature data and very good agreement was obtained. The results show that a strategy of sequential combination of ANN and CRF can provide a more flexible and efficient optimization method for chromatographic separation. © 2009 American Institute of Chemical Engineers AIChE J, 2010     

Minimum Time Heating Cycles for Diffusion‐Controlled Binder Removal from Ceramic Green Bodies

An algorithm based on variational calculus has been developed to predict the minimum time heating cycle (MTHC) for thermal binder removal from a ceramic green body when diffusion, as described by the free volume theory, is the governing mass transport mechanism. The algorithm uses a previously derived analytic solution for the diffusant concentration, which was obtained from the governing reaction–diffusion differential equation. Either a constraint on diffusant concentration or on the equilibrium pressure of diffusant is used to predict the MTHC for both a stationary binder model and a shrinking core binder model. For these four cases, the dependence of the MTHC has been determined on a number of model parameters, including the threshold concentration or pressure, the body size, and the reaction order of the binder degradation kinetics. The algorithm determines two important aspects of the MTHC, namely, the starting temperature of the heating cycle and how temperature varies with time during the cycle. The duration and shape of the temperature‐versus‐time heating schedule, whether increasing, decreasing, or almost constant, depends sensitively on parameters in the model.     

Toward an intelligent approach for predicting surface tension of binary mixtures containing ionic liquids

Knowledge of the surface tension of ionic liquids (ILs) and their related mixtures is of central importance and enables engineers to efficiently design new processes dealing with these fluids on an industrial scale. It’s obvious that experimental determination of surface tension of every conceivable IL and its mixture with other compounds would be a herculean task. Besides, experimental measurements are intrinsically laborious and expensive; therefore, accurate prediction of the property using a reliable technique would be overwhelmingly favorable. To do so, a modeling method based on artificial neural network (ANN) trained by Bayesian regulation back propagation training algorithm (trainbr) has been proposed to predict surface tension of the binary ILs mixtures. A total set of 748 data points of binary surface tension of IL systems within temperature range of 283.1-348.15 K was used to train and test the applied network. The obtained results indicated that the predictive values and experimental data are quite matching, representing reliability of the used ANN model for such purpose. Also, compared with other methods, such as SVM, GA-SVM, GA-LSSVM, CSA-LSSVM, GMDH-PNN and ANN trained with trainlm algorithm the proposed model was better in terms of accuracy.