Modelling the extraction process parameters of amorphous silica-rich rice husk ash using hybrid RSM−BPANN−MOGA optimization technique

Nayak, Puja Priyadarshini; Nandi, Sujosh; Bhunia, Kanishka; Datta, Anindita

doi:10.1016/j.matchemphys.2022.126944

Cited by 6 publications

(3 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The network with the minimum of total MSE of training, validation, and testing subsets was determined as the optimal network architecture, and these models are evaluated for prediction and fitting accuracy with corresponding MSE and R 2 for all the data set following Equations (25) and (26). 24

MSE = \frac{\sum_{i = 1}^{N} {((), Y_{i, e} goodbreak- Y_{i, p})}^{2}}{N}

R^{2} = 1 - \frac{\sum {((), Y_{i, e} goodbreak- Y_{i, p})}^{2}}{\sum {((), Y_{i, e} goodbreak- Y)}^{2}}

where Y i,e is the experimental value of the ith experiment, Y i,p is the corresponding predicted value by the model of the ith experiment, Y is the mean value and N is the number of the experiment. These models were also checked with random data sets to check overlearning or overfitting with more precise predictions.…”

Section: Methodsmentioning

confidence: 99%

“…Ferrández et al 17 also successfully replaced constrained mono‐objective problems with MOO for high‐pressure thermal processes in food treatment, as the latter gave a better, adequate set of parameters with less computation time. Several studies have been dedicated to MOO through GA 16,18,19 and hybrid ANN‐MOGA 20–24 . The utilization of ANN and GA can be efficient for multivariate modeling and optimization in the case of non‐linear variables.…”

Section: Introductionmentioning

confidence: 99%

“…Several studies have been dedicated to MOO through GA 16,18,19 and hybrid ANN-MOGA. [20][21][22][23][24] The utilization of ANN and GA can be efficient for multivariate modeling and optimization in the case of non-linear variables.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Ascorbic acid nanoencapsulation using polyelectrolyte complex formation and optimization using hybrid artificial neural network‐genetic algorithm

Khuntia,

Mitra

2023

JSFA Reports

View full text Add to dashboard Cite

BackgroundsThe polyelectrolyte complex (PEC) formation using chitosan (Cs) and alginate (Alg) has been employed to carry out Vitamin C (VC) nanoencapsulation because of its instability in adverse conditions. This method includes VC nanoencapsulation using physical cross‐linking. The effect of the mixing order as well as the different mass ratios of both the polymer and the VC content, were investigated on the size, pdi, zeta potential, encapsulation efficiency (EE%), and yield%. Hence, considering different mass ratios of both polymer (4:1–1:4) and VC content (10%–30% w/w of Cs) as independent variables, PECs were formed using 0.1% (w/v) of Cs and Alg solution.ResultThe result showed that, for equal mass ratio (1:1) of Cs and Alg, the addition of Alg solution to Cs solution led to lower particle size, while higher particle size was observed with reverse order of addition. However, no significant effect of polymer mass ratios was observed on particle size and NPs stability, while EE varied with VC concentration. Artificial neural network (ANN) was applied to train the experimental parameters Cs: Alg ratios (X1:X2) and VC content (X3) for the output variables. Moreover, process optimization was carried out using multi‐objective genetic algorithm (MOGA) with the goal of lowering particle size and increasing EE while increasing nanoparticle yield.ConclusionThe PEC method effectively encapsulated VC as obtained from higher EE. The Cs: Alg ratio of 4:2.1 and 18% VC content (w/w of Cs) was found optimum for nanoencapsulation. The final ANN‐GA model showed an acceptable agreement with experimental data.

show abstract