Optimization and prediction of safranin-O cationic dye removal from aqueous solution by emulsion liquid membrane (ELM) using artificial neural network-particle swarm optimization (ANN-PSO) hybrid model and response surface methodology (RSM)

Fetimi, Abdelhalim; Dâas, Attef; Benguerba, Yacine; Merouani, Slimane; Hamachi, Mourad; Kebiche-Senhadji, Ounissa; Hamdaoui, Oualid

doi:10.1016/j.jece.2021.105837

Cited by 39 publications

(11 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Input (IL), output (OL), and intermediate or hidden (HL) layers ( Figures S1 and S2 of the Supporting Information) are the three architectural levels of ANNs. 51 , 53 , 54 Neurons (or nodes) are essential computer components used in parallel computing. 53 − 55 Neurons work together to detect input data sets commonly seen.…”

Section: Methodsmentioning

confidence: 99%

“… 51 , 53 , 54 Neurons (or nodes) are essential computer components used in parallel computing. 53 − 55 Neurons work together to detect input data sets commonly seen. 56 f is the transfer function; x j is the neuron’s inputs; and w ij is the link between IL and HL (weights) and the HL’s threshold of j neuron.…”

Section: Methodsmentioning

confidence: 99%

“… 56 f is the transfer function; x j is the neuron’s inputs; and w ij is the link between IL and HL (weights) and the HL’s threshold of j neuron. 50 , 53 Constructing and training neuronal networks has as its primary goal the minimization of the objective function (or fitness function), which in turn leads to better predictions for new input. 52 , 57 The latter compares the output and experimental data sets to determine how well the network operates.…”

Section: Methodsmentioning

confidence: 99%

“…Input (IL), output (OL), and intermediate or hidden (HL) layers (Figures S1 and S2 of the Supporting Information) are the three architectural levels of ANNs. ,, Neurons (or nodes) are essential computer components used in parallel computing. − Neurons work together to detect input data sets commonly seen f is the transfer function; x j is the neuron’s inputs; and w ij is the link between IL and HL (weights) and the HL’s threshold of j neuron. , Constructing and training neuronal networks has as its primary goal the minimization of the objective function (or fitness function), which in turn leads to better predictions for new input. , The latter compares the output and experimental data sets to determine how well the network operates. The fitness (error) function may thus be written as follows: N is the number of experiments, and y i and y o are the calculated and experimental data, respectively …”

Section: Methodsmentioning

confidence: 99%

See 3 more Smart Citations

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H₂O₂/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model

Fetimi

Merouani²,

Khan

et al. 2022

ACS Omega

Self Cite

View full text Add to dashboard Cite

An efficient optimization technique based on a metaheuristic and an artificial neural network (ANN) algorithm has been devised. Particle swarm optimization (PSO) and ANN were used to estimate the removal of two textile dyes from wastewater (reactive green 12, RG12, and toluidine blue, TB) using two unique oxidation processes: Fe(II)/chlorine and H 2 O 2 /periodate. A previous study has revealed that operating conditions substantially influence removal efficiency. Data points were gathered for the experimental studies that developed our ANN-PSO model. The PSO was used to determine the optimum ANN parameter values. Based on the two processes tested (Fe(II)/chlorine and H 2 O 2 /periodate), the proposed hybrid model (ANN-PSO) has been demonstrated to be the most successful in terms of establishing the optimal ANN parameters and brilliantly forecasting data for RG12 and TP elimination yield with the coefficient of determination (R2) topped 0.99 for three distinct ratio data sets.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H₂O₂/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model

Fetimi

Merouani²,

Khan

et al. 2022

ACS Omega

Self Cite

View full text Add to dashboard Cite

show abstract

“…This algorithm is based on the modeling of colonial phenomenon process and is competitive Fig. 3 Structure of neural network with other evolutionary algorithms [34]. In the next Nvar optimization problem, a country is a 1 × Nvar array.…”

Section: Imperialist Competitive Algorithmmentioning

confidence: 99%

Prediction of mechanical properties and hardness of friction stir welding of Al 5083/pure Cu using ANN, ICA and PSO model

et al. 2022

View full text Add to dashboard Cite

The effect of parameters involved in prediction of mechanical properties of Friction Stir Welding have been investigated in this study through modeling. For this purpose, the friction stir welding on both sheets of 5083 Aluminum (Al5083) and pure Copper (Cu) was experimentally conducted at first level. Three factors of Rotation Speed (RPM), Traverse Speed (mm/min) and the Tool’s pin angle was examined. During the tentative test, many experiments were not satisfactorily performed. Thanks to a more rigorous study hypothesis, proper samples were obtained by changing the geometry of different tools. In order to overcome the softness of aluminum sheets and the rise of shoulder diameter, and also to avoid the vertical instability of the sheet, shortening the pin’s length was suggested. In this paper, the Full Factorial method has been employed to evaluate the result of Artificial Neural Networks (ANN), Imperialist Competitive Algorithm, Particle Swarm Optimization, and also the effects of input parameters of the process on output parameters. Moreover, Al5083 and Cu joining sheets were analyzed. Micro-hardness and tensile tests have been based on the process’ input parameters to obtain mechanical properties. The function of ANN model demonstrates that it can estimate the number of mechanical properties with an adequate precision. Using the evaluation factors of mechanical properties and micro-hardness and also a R2 ~ 0.943 analysis, the optimum parameters of Al5083 and Cu joining sheets can be anticipated. The rotation speed of 1150 rpm, traverse speed of 40 mm/min, and pin angle of 2° are the optimum conditions based on the average review of analyses. These optimum conditions led to improve the maximum tensile force up to 15 ~ 21%.

show abstract

Machine learning for membrane design in energy production, gas separation, and water treatment: a review

Osman,

Nasr,

Farghali

et al. 2024

Environ Chem Lett

View full text Add to dashboard Cite

Membrane filtration is a major process used in the energy, gas separation, and water treatment sectors, yet the efficiency of current membranes is limited. Here, we review the use of machine learning to improve membrane efficiency, with emphasis on reverse osmosis, nanofiltration, pervaporation, removal of pollutants, pathogens and nutrients, gas separation of carbon dioxide, oxygen and hydrogen, fuel cells, biodiesel, and biogas purification. We found that the use of machine learning brings substantial improvements in performance and efficiency, leading to specialized membranes with remarkable potential for various applications. This integration offers versatile solutions crucial for addressing global challenges in sustainable development and advancing environmental goals. Membrane gas separation techniques improve carbon capture and purification of industrial gases, aiding in the reduction of carbon dioxide emissions.

show abstract

Optimization and prediction of safranin-O cationic dye removal from aqueous solution by emulsion liquid membrane (ELM) using artificial neural network-particle swarm optimization (ANN-PSO) hybrid model and response surface methodology (RSM)

Cited by 39 publications

References 33 publications

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H₂O₂/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H₂O₂/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model

Prediction of mechanical properties and hardness of friction stir welding of Al 5083/pure Cu using ANN, ICA and PSO model

Machine learning for membrane design in energy production, gas separation, and water treatment: a review

Contact Info

Product

Resources

About

Optimization and prediction of safranin-O cationic dye removal from aqueous solution by emulsion liquid membrane (ELM) using artificial neural network-particle swarm optimization (ANN-PSO) hybrid model and response surface methodology (RSM)

Cited by 39 publications

References 33 publications

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H2O2/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H2O2/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model

Prediction of mechanical properties and hardness of friction stir welding of Al 5083/pure Cu using ANN, ICA and PSO model

Machine learning for membrane design in energy production, gas separation, and water treatment: a review

Contact Info

Product

Resources

About

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H₂O₂/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model

Modeling of Textile Dye Removal from Wastewater Using Innovative Oxidation Technologies (Fe(II)/Chlorine and H₂O₂/Periodate Processes): Artificial Neural Network-Particle Swarm Optimization Hybrid Model