Hybrid Artificial Neural Network-genetic Algorithm Research Articles

Modeling and optimization of activated sludge bulking for a real wastewater treatment plant using hybrid artificial neural networks-genetic algorithm approach

Prediction of sludge bulking is a matter of growing importance around the world. Sludge volume index (SVI) should be monitored to predict sludge bulking for a wastewater treatment plant. This study was an effort to develop hybrid artificial neural network-genetic algorithm models (MLPANN-GA and RBFANN-GA) to accurately predict SVI. Operating parameters, including MLVSS, pH, DO, temperature, TSS, COD and total nitrogen were the inputs of neural networks. Genetic algorithm was utilized in order to optimize weights and thresholds of the MLPANN and RFBANN models. Training procedures for SVI estimation were successful for both the MLPANN-GA and RBFANN-GA models. The training and validation models showed an almost perfect match between experimental and predicted values of SVI. The results indicated that with low experimental values of input data to train ANNs, the MLPANN-GA compared with the RBFANN-GA is more accurate due to higher coefficient of determination (R2) and lower root mean squared error (RMSE) values. The values of RMSE and R2 for the optimal models approached 0 and 1, respectively. The mean average error for the ANN models did not exceed 3% of the input values of the measured SVI. The GA increased the accuracy of all the MLPANN and RBFANN models.

Developing a hybrid artificial neural network-genetic algorithm model to predict resilient modulus of polypropylene/polyester fiber-reinforced asphalt concrete

Up to now various kinds of fibers are used to improve the hot mix asphalt (HMA) performance, but a few works have been undertaken on the hybrid fiber-reinforced HMA. Therefore, in this paper, the resilient modulus of the modified HMA samples using polypropylene and polyester fibers (hybrid and single modes) was evaluated and modeled by regression method and artificial neural network (ANN). As ANN includes different parameters such as the number of neurons in hidden layer influenced on the prediction accuracy, genetic algorithm (GA) was used to optimize the ANN parameters. Also, GA parameters were optimized using the trial and error method such as the population size. The obtained results indicated that the optimized ANN with two hidden layers and two neurons in each hidden layer can predict the resilient modulus of fiber-reinforced HMA with high accuracy (correlation coefficient: .96).

Prediction of the extraction yield using artificial neural network and response surface methodology: ultrasound-assisted extraction from Achillea berbresteinii L.

This study investigates the extraction efficiency of phenolic compounds from Achillea berbresteinii by ultrasound-assisted extraction (UAE) method. Meanwhile, to predict the phenolic compound extraction yield, artificial neural network-genetic algorithm (ANN-GA) and response surface methodology (RSM) were compared. The results indicated that UAE method could significantly improve the extraction yield in comparison to conventional method. Optimised processing conditions were 35 °C, 6.3, 20% and 35 min as temperature, pH, solvent to sample ratio and extraction time, respectively. On the other hand, hybrid ANN-GA was employed to estimate the phenolic compound extraction yield. The results revealed the better capability of this method in comparison to RSM. Optimised network contained 8 and 3 neurons in first and second hidden layers, respectively. This configuration could estimate phenolic compound extraction yield with high correlation coefficient (0.94). Finally, A. berbresteinii can be considered as an excellent potential source of phenolic compounds and ANN-GA as a successful applied method for the prediction of the phenolic compound extraction yield.

APPLICATION OF COMPUTATIONAL INTELLIGENCE METHOD IN PETROLEUM GEOMECHANICS CHARACTERIZATION

Petroleum geomechanics characterization refers to the process of quantitatively assigning geomechanical parameters using all available field data. In this paper, an attempt is made to develop a computational intelligence method that integrates genetic algorithm (GA) and artificial neural network (ANN). Through this method, these geomechanical parameters such as horizontal in situ stresses, fracture aperture and joint spacing are determined based on the borehole displacements during drilling well. In the hybrid ANN–GA model, GA can automatically identify geomechanical parameters as neural inputs from monitored borehole displacements, and the ANN is trained to learn the complex relationship between geomechanical parameters and target borehole displacements. Data from numerical experiments on petroleum wells are used in verification of the proposed computational intelligence method for geomechanics characterization. The study of numerical experiment illustrates that the proposed computational intelligence method has the ability to generate reliable results.

Production of Engineered Fabrics Using Artificial Neural Network–Genetic Algorithm Hybrid Model

The process of fabric engineering which is generally practised in most of the textile mills is very complicated, repetitive, tedious and time consuming. To eliminate this trial and error approach, a new approach of fabric engineering has been attempted in this work. Data sets of construction parameters [comprising of ends per inch, picks per inch, warp count and weft count] and three fabric properties (namely drape coefficient, air permeability and thermal resistance) of 25 handloom cotton fabrics have been used. The weights and biases of three artificial neural network (ANN) models developed for the prediction of drape coefficient, air permeability and thermal resistance were used to formulate the fitness or objective function and constraints of the optimization problem. The optimization problem was solved using genetic algorithm (GA). In both the fabrics which were attempted for engineering, the target and simulated fabric properties were very close. The GA was able to search the optimum set of fabric construction parameters with reasonably good accuracy except in case of EPI. However, the overall result is encouraging and can be improved further by using larger data sets of handloom fabrics by hybrid ANN–GA model.

Application of a hybrid artificial neural network–genetic algorithm approach to optimize the lead ions removal from aqueous solutions using intercalated tartrate-Mg–Al layered double hydroxides

The aim of this work was to develop an eco-friendly adsorbent to remove the lead ions from aqueous solutions. The present study uses the artificial neural network (ANN) and genetic algorithm (GA) as tools for simulation and optimization of the lead ions removal from aqueous solutions using intercalated tartrate-Mg–Al layered double hydroxides as an adsorbent. A multi-layer ANN model trained with Levenberg–Marquardt (LM) algorithm was incorporated for developing a predictive model. The input variables of the model were time, solution pH, adsorbent dosage, and lead ion concentration, while the percentage of lead ions removal was the output. The performance of the ANN model was determined and showed the efficiency of the model to predict the percentage of lead ions removal accurately with determination coefficient (R2) of 0.99986. The developed ANN model was then used with GA to optimize the percentage of lead ions removal from aqueous solution. The optimized conditions were obtained as follows: 11.8h, 4.5, 0.054g and 10.0mg/L for the time, solution pH, adsorbent dosage, and lead ion concentration, respectively. Under these optimum conditions, the predicted and actual percentages of lead ions removal were 101.2 and 98.7%, respectively.

Yarn engineering using hybrid artificial neural network-genetic algorithm model

This work aims to manufacture cotton yarns with requisite quality by choice of suitable raw materials for a given spinning system. To fulfill this aim, a hybrid model based on Artificial Neural Network (ANN) and Genetic Algorithm (GA) has been developed which captures both the high prediction power of ANN and global solution searching ability of GA. In an attempt to achieve a yarn having predefined tenacity and evenness, a constrained optimization problem is formulated with the ANN input-output relation between fibre and yarn properties. GA has been used to solve the optimization problem by searching the best combination of fibre properties that can translate into reality a yarn with the desired quality. The model is capable in identifying the set of fibre properties that gives requisite yarn quality with reasonable degree of accuracy.

PREDICTION OF THE CHLOROBUTYL RUBBER/NATURAL RUBBER BLEND PROPERTIES USING A GENETIC ALGORITHM AND ARTIFICIAL NEURAL NETWORK

ABSTRACT In this study, an orthogonal experimental design (OED) was used to predict and explain the effect of the rubber ratio and the carbon black (CB) and processing oil concentration on the mechanical and gas barrier properties of chlorobutyl rubber (CIIR)/natural rubber (NR) blends. The OED method was adopted for its advantage that it can select representative test points from comprehensive tests and so minimize the number of experiments required to be performed. The hybrid genetic algorithm-artificial neural network technique was applied to obtain an optimal formulation of CIIR/NR, CB, and oil compound with suitable mechanical properties and a good gas barrier property. The optimal composition found exhibited improved mechanical and gas barrier properties, which were found to depend mainly on the rubber ratio and the CB and oil concentrations. With respect to industrial applications, the optimum rubber ratio of CIIR/NR was found to be 87.5 phr/12.5 phr, and the optimum concentrations of CB and oil were 66.25 phr and 10.25 phr, respectively.

Bulk modulus prediction of austenitic stainless steel using a hybrid GA–ANN as a data mining tools

In the current paper, the hybrid model based on genetic algorithm (GA) and artificial neural network (ANN) was used as a data mining tool to synthesize the optimal concentration of manganese (Mn). The aim is to achieve the optimal bulk modulus of FeCrNiMn austenitic stainless steel alloy. An ANN model has been developed to analyze and simulate the correlation between the elastic properties and chemical composition. The ANN training has been carried out upon three inputs, namely (Cr, Ni, and Mn), with an alloy weight percentage each, while the bulk modulus is the output target. The fitness function of GA was obtained from trained ANN model. The Mn concentration value has been obtained by the GA–ANN algorithm corresponding at the maximum bulk modulus. More ever, the given result through the GA–ANN was compared to the obtained from quantum mechanical simulation, based on the first principal calculations implemented in the Vienna Ab-initio Simulation Package (VASP). It was found that the relative error is within 2.89%. The results averred that the data mining tool based on the combination of GA–ANN is a useful, efficient, strong and adequate.

Hybrid ANN–GA model for predicting turbidity and chlorophyll-a concentrations

The transport and fate of admixtures at coastal zones are driven, or at least modulated, by currents. In particular, in tide-dominated areas due to higher near-bottom shear stress at strong currents, sediment concentration and turbidity are expected to be at maximum during spring tide, while algal growth rate likely is peaking up at slack currents during neap tide. Varying weather and atmospheric conditions might modulate the said dependencies, but the water quality pattern still is expected to follow the dominant tidal cycle. As tidal cycling could be predicted well ahead, there is a possibility to use water quality and hydrodynamic high-resolution data to learn past dependencies, and then use tidal hydrodynamic model for nowcasting and forecasting of selected water quality parameters.This paper develops data driven models for nowcasting and forecasting turbidity and chlorophyll-a using Artificial Neural Network (ANN) combined with Genetic Algorithm (GA). The use of GA aims to automate and enhance ANN designing process. The training of the ANN model is done by constructing input–output mapping, where hydrodynamic parameters act as an input for the network, while turbidity and chlorophyll-a are the corresponding outputs (desired target). Afterward, the prediction is carried out only by employing computed water surface elevation as an input for the trained ANN model. The proposed data driven model has successfully revealed complex relationships and utilized its experiential knowledge acquired from the training process for facilitating the subsequent use of the data driven model to yield an accurate prediction.

Artificial intelligence based modeling and optimization of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) production process by using Azohydromonas lata MTCC 2311 from cane molasses supplemented with volatile fatty acids: A genetic algorithm paradigm

The present work describes the optimization of medium variables for the production of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) [P(3HB-co-3HV)] by Azohydromonas lata MTCC 2311 using cane molasses supplemented with propionic acid. Genetic algorithm (GA) has been used for the optimization of P(3HB-co-3HV) production through the simulation of artificial neural network (ANN) and response surface methodology (RSM). The predictions by ANN are better than those of RSM and in good agreement with experimental findings. The highest P(3HB-co-3HV) concentration and 3HV content have been reported as 7.35g/l and 16.84mol%, respectively by hybrid ANN–GA. Upon validation, 7.20g/l and 16.30mol% of P(3HB-co-3HV) concentration and 3HV content have been found in the shake flask, whereas 6.70g/l and 16.35mol%, have been observed in a 3l bioreactor, respectively. The specific growth rate and P(3HB-co-3HV) accumulation rate of 0.29 per h and 0.16g/lh determined with cane molasses are comparable to those observed on pure substrates.

Prediction of mechanical properties of waste polypropylene/waste ground rubber tire powder blends using artificial neural networks

Recycling represents a valid alternative to the disposal of post-consumer materials if it is possible to obtain new materials with good properties. In this work, waste polypropylene (WPP)/waste ground rubber tire (WGRT) powder blends were studied with respect to the effect of bitumen and maleic anhydride-grafted styrene–ethylene–butylene–styrene (SEBS-g-MA) content by using the design of experiments (DOE) approach, whereby the effect of the four polymers content on the final mechanical properties were predicted. Uniform design method was especially adopted for its advantages. Optimization was done using hybrid artificial neural network–genetic algorithm (ANN–GA) technique. The results indicated that the blends show fairly good ductibility provided that it had a relatively higher concentration of bitumen and SEBS-g-MA under the studied condition. A quantitative relationship was presented between the material concentration and the mechanical properties as a set of contour plots, which were confirmed experimentally by testing the optimum ratio.

Prediction of mechanical properties of polypropylene/waste ground rubber tire powder treated by bitumen composites via uniform design and artificial neural networks

Polypropylene (PP)/waste ground rubber tire powder (WGRT) composites were studied with respect to the effect of bitumen and maleic anhydride-grafted styrene–ethylene–butylene–styrene (SEBS-g-MA) content by using the design of experiments (DOE) approach, whereby the effect of the four polymers content on the final mechanical properties were predicted. Uniform design method was especially adopted for its advantages. Optimization was done using hybrid artificial neural network–genetic algorithm (ANN–GA) technique. The results indicated that the composites showed fairly good ductibility provided that it had a relatively higher concentration of bitumen and SEBS-g-MA under the studied condition. A quantitative relationship was presented between the material concentration and the mechanical properties as a set of contour plots, which were confirmed experimentally by testing the optimum ratio.

On-capillary derivatization using a hybrid artificial neural network-genetic algorithm approach

The first reported hybrid artificial neural network-genetic algorithm (ANN-GA) approach for the optimization of on-capillary dipeptide derivatization is presented. More specifically, genetic optimization proved valuable in the determination of effective network structure with three defined parameter inputs: (i) phthalic anhydride injection volume, (ii) time of injection, and (iii) voltage, for the maximum conversion of the dipeptide D-Ala-D-Ala by phthalic anhydride. Results obtained from the hybrid approach proved superior to an ANN model without GA optimization in terms of training data and predictive ability. The model developed will likely prove useful for the analysis of other organic-based reaction systems.

Comparison of artificial neural network (ANN) and response surface methodology (RSM) in fermentation media optimization: Case study of fermentative production of scleroglucan

Response surface methodology (RSM) is the most preferred method for fermentation media optimization so far. In last two decades, artificial neural network-genetic algorithm (ANN-GA) has come up as one of the most efficient method for empirical modeling and optimization, especially for non-linear systems. This paper presents the comparative studies between ANN-GA and RSM in fermentation media optimization. Fermentative production of biopolymer scleroglucan has been chosen as case study. The yield of scleroglucan was modeled and optimized as a function of four independent variables (media components) using ANN-GA and RSM. The optimized media produced 16.22 ± 0.44 g/l scleroglucan as compared to 7.8 ± 0.54 g/l with unoptimized medium. Two methodologies were compared for their modeling, sensitivity analysis and optimization abilities. The predictive and generalization ability of both ANN and RSM were compared using separate dataset of 17 experiments from earlier published work. The average % error for ANN and RSM models were 6.5 and 20 and the CC was 0.89 and 0.99, respectively, indicating the superiority of ANN in capturing the non-linear behavior of the system. The sensitivity analysis performed by both methods has given comparative results. The prediction error in optimum yield by hybrid ANN-GA and RSM were 2% and 8%, respectively.

Prediction and Optimization of Mechanical Properties of Polypropylene/Waste Tire Powder Blends using a Hybrid Artificial Neural Network-Genetic Algorithm (GA-ANN)

Blends of Polypropylene (PP) and waste ground rubber tire powder are studied with respect to the effect of ethylene—propylene—diene monomer (EPDM) and polypropylene grafted maleic anhydride (PP-g-MA) compatibilizer content by using the Design of Experiments methodology, whereby the effect of the four polymers content on the final mechanical properties are predicted. Uniform design method is especially adopted for its advantages. Optimization is done using hybrid Artificial Neural Network-Genetic Algorithm technique. A rubber formulary with respect to the four ingredients are optimized having maximum tensile strength and then compared with a blend predicted to have maximum elongation at break. It is concluded that the blends show fairly good properties provided that it has a relatively higher concentration of PP-g-MA and EPDM content. SEM investigations also corroborates with the observed mechanical properties. A quantitative relationship is then shown between the material concentration and the mechanical properties as a set of contour plots, which are then tested and confirmed experimentally to conform to the optimum blend ratio.

Modelling and optimization of catalytic–dielectric barrier discharge plasma reactor for methane and carbon dioxide conversion using hybrid artificial neural network—genetic algorithm technique

A hybrid artificial neural network-genetic algorithm (ANN-GA) was developed to model, simulate and optimize the catalytic–dielectric barrier discharge plasma reactor. Effects of CH 4 / CO 2 feed ratio, total feed flow rate, discharge voltage and reactor wall temperature on the performance of the reactor was investigated by the ANN-based model simulation. Pareto optimal solutions and the corresponding optimal operating parameter range based on multi-objectives can be suggested for two cases, i.e., simultaneous maximization of CH 4 conversion and C 2 + selectivity (Case 1), and H 2 selectivity and H 2 / CO ratio (Case 2). It can be concluded that the hybrid catalytic–dielectric barrier discharge plasma reactor is potential for co-generation of synthesis gas and higher hydrocarbons from methane and carbon dioxide and performed better than the conventional fixed-bed reactor with respect to CH 4 conversion, C 2 + yield and H 2 selectivity.