Artificial neural networks: an efficient tool for modelling and optimization of biofuel production (a mini review)

Figures & data

Figure 1. General topology of a multilayer structure of an artificial neural network.

Table 1. Summary of modelling studies using ANN for biohydrogen production.

Input parameters	Output parameters	Type of ANN	ANN structure	R^2 value	References
pH, glucose:xylose ratio, inoculum size, inoculum age	Cumulative H2	FFBPNN	4-10-1	0.99	[29]
T °C, pH, So (glucose)	HY	FFBPNN	3-4-1	–	[9]
T °C, pH, So (glucose)	SE(%), HPR, HY	FFBPNN	3-5-1	0.98, 0.99, 0.98	[10]
So (molasses), inoculum %, T °C	Cumulative H2	FFBPNN	4-(6-10)-1	0.91	[24]
ORP, pH, dissolved CO2	HPR	FFBPNN	–	0.96	[45]
HRT, So (sucrose), ORP, pH, recycle ratio, alkalinity	HPR	FFBPNN	12-20-1	0.80	[46]
OLR, ORP, pH, alkalinity	HPR	FFBPNN	4-3-1	–	[33]
pH, So (glucose, sucrose and thin sillage), Xo, T °C, time	HPR	FFBPNN	5-6-4-1	0.98	[47]
OLR, pH, VSS yield	HPR	FFBPNN	3-8-4-1	0.85	[48]
OLR, HRT, influent alkalinity	HY, HPR, TOCeff, products conc.	FFBPNN	–	–	[49]
So (xylose), peptone concentration, pH	HY	FFBPNN	3-8-1	0.99	[50]
T °C, So (glucose), pH	HY	FFBPNN	3-6-4-2-1	0.98	[51]
Acidity, pH, So (glucose), HRT	HY	FFBPNN	4-12-4-1	0.99	[52]
Inoculum type, substrate type, So (glucose, sucrose, xylose), pH, T °C	HY	FFBPNN	5-7-7-1	0.90, 0.46	[53]

Note: ORP, oxidation--reduction potential; HPR, hydrogen production; HRT, hydraulic retention time; S_o, initial substrate-concentration; SE%, substrate degradation efficiency; X_o, initial biomass concentration; T °C, temperature; SE, substrate degradation efficiency; OLR, organic loading rate; TOC_eff, effluent total organic carbons; VSS, volatile suspended solids; FFBPNN, feed forward back-propagation neural network; HY, hydrogen yield; R², coefficient of determination.

Table 2. Summary of modelling studies using ANN for biogas production.

Input parameters	Output parameters	Type of ANN	ANN structure	R^2 value	References
Flow rate, volumetric load, CODin, TSSin	CODout, TSSout, biogas production	GRNN RBF FFBPNN	–	–	[54]
OLR, VFA, influent–effluent alkalinity, influent–effluent pH, T °C	Biogas production	FFBPNN	–	0.93	[55]
Sludge concentrations	Methane production	FFBPNN	5-7-1	0.99	[56]
Co-substrates concentration	Biogas production	FFBPNN	5-2-1	–	[11]
Leachate (pH, alkalinity, COD, sulphate, conductivity, chloride, waste T °C) and refuse age	Methane fraction (%) in biogas	FFBPNN	–	0.96	[57]
Peak current, pre-peak slope	COD removal efficiency (%), methane production	FFBPNN	2-3-1	–	[58]
T °C, pH, TS, TVS	Biogas yield	FFBPNN	–	0.87	[12]
H2S:S LR, H2S in biogas, total sulphides, pH, OLR	H2S and NH3 in biogas	FFBPNN	4-3-1	0.91, 0.83	[59]

Note: COD_in, chemical oxygen demand (initial); TSS_in, total suspended solids (initial); COD_out, chemical oxygen demand (final); TSS_out, total suspended solids (final); OLR, organic loading rate; VFA, volatile fatty acids; T °C, temperature; TS, total solid; TVS, total volatile solid; H₂S:S, hydrogen sulphide-to-sulphur ratio; GRNN, generalized regression neural networks; RBF, radial basis function-based neural network; FFBPNN, feed forward back-propagation neural network; LR, loading rate; R², coefficient of determination.

Table 3. Summary of modelling studies using ANN for microbial fuel cell (MFC) technology.

Input parameters	Output parameters	Type of ANN	ANN structure	R^2 value	References
pH, BOD, COD, TSS	Current generation	FFBPNN	4-4-1	–	[60]
T °C, pH, electron acceptor concentration	Power density, current density	FFBPNN	3-3-1	0.98	[61]
T °C, ferrous sulphate concentration	Voltage	FFBPNN	3-9-1	0.90	[62]
pH, T °C, So, substrate type, applied voltage, MEC configuration	HY	FFBPNN	6-(6, 8, 11, 12, 14)-1	0.90, 0.81, 0.85, 0.70, 0.80	[15]

Note: BOD, biochemical oxygen demand; COD, chemical oxygen demand; TSS, total suspended solids; T °C, temperature; S_o, initial substrate concentration; MEC, microbial electrolysis cell; HY, hydrogen yield; FFBPNN, feed forward back-propagation neural network; R², coefficient of determination.

Table 4. Summary of modelling studies using ANN for biodiesel production.

Input parameters	Output parameters	Type of ANN	ANN structure	R^2 value	References
So (glucose), yeast extract concentration, sodium nitrate concentration	Lipid productivity, biomass concentration	FFBPNN	3-10-1	0.99	[14]
HRT	Biomass concentration	–	–	0.95	[63]
pH	Biomass concentration	–	1-2-1	–	[64]
So (glucose)	Biomass concentration	HNN	1-3-1	–	[65]

Note: S_o, initial substrate concentration; FFBPNN, feed forward back-propagation neural network; HNN, hybrid neural network; R², coefficient of determination; HRT, hydraulic retention time.

Table 5. Summary of modelling studies using ANN for bioethanol production.

Input parameter	Output parameter	Type of ANN	ANN structure	R^2 value	References
So (molasses), live yeast cells, dead yeast cells	Bioethanol production	–	–	0.93	[4]
So (breadfruit starch hydrolysate), HRT, pH	Bioethanol production	FFBPNN	3-3-1	1	[66]
pH, So (glucose), T °C	Biomass yield, bioethanol concentration	FFBPNN	3-6-2	0.99, 0.99	[67]
HRT, So (sugar beet wastewater), substrate type	Bioethanol concentration	FFBPNN	3-12-1	0.99	[68]

Note: S_o, initial substrate concentration; HRT, hydraulic retention time; FFBPNN, feed forward back-propagation neural network; R², coefficient of determination; T °C, temperature.

Figure 2. Feed forward back-propagation training flowchart for artificial neural networks.

Table 6. Summary of comparative modelling studies using ANN and RSM for biofuel production.

Input parameters	Output parameters	ANN R^2	RSM R^2	ANN prediction error (%)	RSM prediction error (%)	References
So (molasses), inoculum %, T °C	Cumulative H2	0.91	0.75	15.12	119.08	[24]
T °C, pH, So (glucose)	HY	–	–	7.70	16.60	[9]
So (glucose), yeast extract concentration and sodium nitrate concentration	Lipid productivity, biomass concentration	0.99	0.99	4.18	5.72	[14]
So (xylose), peptone concentration, pH	HY	0.99	0.96	3	13	[50]
So (breadfruit starch hydrolysate), HRT and pH	Bioethanol production	1	0.99	0.24	3.41	[66]
pH, T °C, So (glucose)	Biomass yield, bioethanol concentration	0.99	0.99	1.90	2.57	[67]

Note: S_o, initial substrate concentration; T °C, temperature; HY, hydrogen yield; HRT, hydraulic retention time; R₂, coefficient of determination.

Prakasham RS, Sathish T, Brahmaiah P. Imperative role of neural networks coupled genetic algorithm on optimization of biohydrogen yield. Int J Hydrog Energy. 2011;36:4332–4339.

 Web of Science ®Google Scholar

Wang JL, Wan W. Optimization of fermentative hydrogen production process using genetic algorithm based on neural network and response surface methodology. Int J Hydrog Energy. 2009;34:255–261.

 Web of Science ®Google Scholar

Wang JL, Wan W. Application of desirability function based on neural network for optimizing biohydrogen production process. Int J Hydrog Energy. 2009;34:1253–1259.

 Web of Science ®Google Scholar

Whiteman JK, Gueguim Kana EB. Comparative assessment of the Artificial Neural Network and Response Surface Modelling efficiencies for biohydrogen production on sugar cane molasses. Bioenergy Res. 2013;7(1):295–305.

 Web of Science ®Google Scholar

Rosales-Colunga LM, Garcia RG, Rodriguez AD. Estimation of hydrogen production in genetically modified E. coli fermentations using an artificial neural network. Int J Hydrog Energy. 2010;35:13186–13192.

 Web of Science ®Google Scholar

Nikhil BO, Visa A, Lin CY, et al. An artificial neural network based model for predicting H2 production rates in a sucrose-based bioreactor system. Int J Chem Mol Nucl Mater Metall Eng. 2008;37:20–25.

 Google Scholar

Shi Y, Gai G, Zhao X, et al. Back propagation neural network (BPNN) simulation model and influence of operational parameters on hydrogen bio-production through integrative biological reactor (IBR) treating wastewater. Proceedings of the 4th International Conference on Bioinformatics and Biomedical Engineering (ICBBE); 2010 Jun 18–20; Chengdu, China. Piscataway (NJ): IEEE; 2010.

 Google Scholar

Nasr N, Hafez H, El Naggar MH, et al. Application of artificial neural networks for biohydrogen production. Int J Hydrog Energy. 2013;38:3189–3195.

 Web of Science ®Google Scholar

Nasr M, Tawfik A, Ookawara S, et al. Prediction of hydrogen production from starch wastewater using Artificial Neural Networks. Int Water Technol J. 2014;4(1):36–44.

 Google Scholar

Mu Y, Yu HQ. Simulation of biological hydrogen production in a UASB reactor using neural network and genetic algorithm. Int J Hydrog Energy. 2007;32:3308–3314.

 Web of Science ®Google Scholar

Karthic P, Joseph S, Arun N, et al. Optimization of biohydrogen production by Enterobacter species using artificial neural network and response surface methodology. J Renew Sust Energy [Internet]. 2013 [cited 2016 Jun 20];5(3):033104. Available from : http://scitation.aip.org/content/aip/journal/jrse/5/3/10.1063/1.4803746

 Web of Science ®Google Scholar

Alalayah WM, Alhamed Y, Al-Zahrani AA, et al. Merits of utilizing an artificial neural network as a prediction model for bio-hydrogen production. Revista Chimi (Bucharest). 2014;65(4):458–465.

 Web of Science ®Google Scholar

Mullai P, Yogeswari MK, Sridevi K, et al. Artificial neural network (ANN) modeling for hydrogen production in continuous anaerobic sludge blanket filters (ASBF). Singap J Sci Res. 2013;5(1):1–7.

 Google Scholar

Sewsynker Y, Gueguim Kana EB. Intelligent models to predict hydrogen yield in dark microbial fermentations using existing knowledge. Int J Hydrog Energy. 2016;41(30):12929–12940.

 Web of Science ®Google Scholar

Elnekave M, Celik SO, Tatlier M, et al. Artificial Neural Network predictions of upflow anaerobic sludge blanket (UASB) reactor performance in the treatment of citrus juice wastewater. Pol J Environ Stud. 2012;21(1):49–56.

 Web of Science ®Google Scholar

Kanat G, Saral A. Estimation of biogas production rate in a thermophilic UASB reactor using Artificial Neural Networks. Environ Model Assess. 2009;14:607–614.

 Web of Science ®Google Scholar

Mahanty B, Zafar M, Park HS. Characterization of co-digestion of industrial sludges for biogas production by artificial neural network and statistical regression models. Environ Technol. 2013;34:2145–2153.

 PubMed Web of Science ®Google Scholar

Gueguim Kana EB, Oloke JK, Lateef A, et al. Modelling and optimization of biogas production on saw dust and other co-substrates using artificial neural network and genetic algorithm. Renew Energy. 2012;46:276–281.

 Web of Science ®Google Scholar

Ozkaya B, Demir A, Bilgili MS. Neural network prediction model for the methane fraction in biogas from field-scale landfill bioreactors. Environ Modelling Softw. 2007;22:815–822.

 Web of Science ®Google Scholar

Harper WF, Jr. Taewoo Y. Using electronic signals and neural networks to monitor the performance of an anaerobic bioreactor. Int J Water Res Environ Eng. 2013;5(9):521–532.

 Google Scholar

Abu-Qdais H, Bani-Hani K, Shatnawi N. Modeling and optimization of biogas production from a waste digester using artificial neural network and genetic algorithm. Resour Conserv Recycl. 2012;54:359–363.

 Web of Science ®Google Scholar

Strik D, Domnanovich AM, Zani L, et al. Prediction of trace compounds in biogas from anaerobic digestion using the Matlab neural network toolbox. Environ Modelling Softw. 2005;20:803–810.

 Web of Science ®Google Scholar

Tardast A, Rahimnejad M, Najafpour G, et al. Prediction of bioelectricity production by neural network. J Biotechnol Pharm Res. 2012;3(3):62–68.

 Google Scholar

Tardast A, Rahimnejad M, Najafpour G, et al. Use of artificial neural network for the prediction of bioelectricity production in a membrane less microbial fuel cell. Fuel. 2014;117:697–703.

 Web of Science ®Google Scholar

Garg A, Vijayaraghavan V, Mahapatra SS, et al. Performance evaluation of microbial fuel cell by artificial intelligence methods. Expert Syst Appl. 2014;41:1389–1399.

 Web of Science ®Google Scholar

Sewsynker Y, Gueguim Kana EB, Lateef A. Modelling of biohydrogen generation in microbial electrolysis cells (MECs) using a committee of artificial neural networks (ANNs). Biotechnol Biotechnol Equip. 2015;29(6):1208–1215.

 Web of Science ®Google Scholar

Mohamed MS, Tan JS, Mohamad R, et al. Comparative analyses of response surface methodology and artificial neural network on medium optimization for Tetraselmis sp. FTC209 grown under mixotrophic condition. Sci World J. 2013;2013:1–14. Article ID 948940.

 Web of Science ®Google Scholar

Furlong VB, Pereira-Filho RD, Margarites AC, et al. Estimating microalgae Synechococcus nidulans daily biomass concentration using neuro-fuzzy network. Ciênc Tecnol Aliment. 2013;33(1):142–147.

 Google Scholar

Sarkar N, Ghosh SK, Bannerjee S, et al. Bioethanol production from agricultural wastes: an overview. Renew Energy. 2013;37:19–27.

 Web of Science ®Google Scholar

Wu Z, Shi X. Optimization for high-density cultivation of heterotrophic Chlorella based on a hybrid neural network model. Lett Appl Microbiol. 2006;44:13–18.

 Web of Science ®Google Scholar

Ahmadian-Moghadam H, Elegado FB, Nayve R. Prediction of ethanol concentration in biofuel production using artificial neural networks. Am J Model Optim. 2013;1(3):31–35.

 Google Scholar

Betiku E, Taiwo AE. Modeling and optimization of bioethanol production from breadfruit starch hydrolyzate vis-a-vis response surface methodology and artificial neural network. Renew Energy. 2015;74:87–94.

 Web of Science ®Google Scholar

Esfahanian M, Nikzad M, Najafpour G, et al. Modeling and optimization of ethanol fermentation using Saccharomyces cerevisiae: response surface methodology and artificial neural network. Chem Ind Chem Eng Q. 2013;19(2):241–252.

 Web of Science ®Google Scholar

Grahovac J, Jokić A, Dodić J, et al. Modelling and prediction of bioethanol production from intermediates and byproduct of sugar beet processing using neural networks. Renew Energy. 2016;85:953–958.

 Web of Science ®Google Scholar