Application of Artificial Intelligence in the Prediction of Thermal Properties of Biomass

Olatunji, Obafemi O.; Akinlabi, Stephen A.; Madushele, Nkosinathi

doi:10.1007/978-3-030-38032-8_4

Cited by 6 publications

(5 citation statements)

References 129 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various characteristics such as proximate composition, elemental composition, higher heating value (HHV), and gross heating value (GHV) are requirements prior to process design and development for energy generation or other value‐added products. Determining these properties through experimental methods requires money, time, effort, and sophisticated instruments 36 . However, in recent years, the estimation of biomass properties using different modeling tools has been reported by several researchers.…”

Section: Models For Prediction Of Biomass Propertiesmentioning

confidence: 99%

See 1 more Smart Citation

Artificial neural networks in valorization process modeling of lignocellulosic biomass

Pradhan

Jaiswal

2022

Biofuels Bioprod Bioref

View full text Add to dashboard Cite

Value-added products such as biofuels, chemicals, enzymes, and many others can be prepared from lignocellulosic biomass (LCB). To achieve high yields of these value-added products, powerful tools such as artificial neural networks (ANN) and adaptive neuro-fuzzy inference systems (ANFIS) can be utilized during process development. In this article, we have therefore reviewed the recent application of ANN and ANFIS in modeling LCB valorization processes. Studies have shown the high predictive capability of both ANN and ANFIS for a range of different processes such as pre-treatment processes (microwave-assisted, organosolv-, ultrasound-assisted pre-treatment and many others), thermal processes (pyrolysis and gasification), enzymatic hydrolysis, and fermentation processes. These tools have also shown outstanding accuracy in predicting elemental composition and thermal characteristics of biomass by using only the proximate composition of LCB as the input information. In combination with evolutionary algorithms like genetic algorithm, particle swarm optimization or ant colony optimization, the ANN and ANFIS tools have shown excellent results in obtaining operational conditions for the efficient production of bioethanol, biogas, organic acids, lignin, and enzymes. However, there are only limited reports of the application of ANN and ANFIS in enzyme, organic acid and lignin production. Further research is therefore required to assess the suitability of using these tools in process development for the production of lignin, enzymes, and organic acids.

show abstract

Section: Models For Prediction Of Biomass Propertiesmentioning

confidence: 99%

“…Determining these properties through experimental methods requires money, time, effort, and sophisticated instruments. 36 However, in recent years, the estimation of biomass properties using different modeling tools has been reported by several researchers.…”

Section: Models For Prediction Of Biomass Propertiesmentioning

confidence: 99%

Artificial neural networks in valorization process modeling of lignocellulosic biomass

Pradhan

Jaiswal

2022

Biofuels Bioprod Bioref

View full text Add to dashboard Cite

show abstract

“…The input data can result from the proximate and ultimate analysis and process parameters. These results can be considered together or separately [52,53]. In this work, the regression modeling was examined for four models based only on pyrolysis process parameters (temperature and residence time); proximate and ultimate data were not needed.…”

Section: Importance Of Mathematical Approaches For Fuel Properties Prmentioning

confidence: 99%

“…The input data can result from the proximate and ultimate analysis and process parameters. These results can be considered together or separately [ 52 , 53 ].…”

Section: Introductionmentioning

confidence: 99%

The Prediction of Calorific Value of Carbonized Solid Fuel Produced from Refuse-Derived Fuel in the Low-Temperature Pyrolysis in CO2

et al. 2020

View full text Add to dashboard Cite

The decrease in the calorific value of refuse-derived fuel (RDF) is an unintended outcome of the progress made toward more sustainable waste management. Plastics and paper separation and recycling leads to the overall decrease in waste’s calorific value, further limiting its applicability for thermal treatment. Pyrolysis has been proposed to densify energy in RDF and generate carbonized solid fuel (CSF). The challenge is that the feedstock composition of RDF is variable and site-specific. Therefore, the optimal pyrolysis conditions have to be established every time, depending on feedstock composition. In this research, we developed a model to predict the higher heating value (HHV) of the RDF composed of eight morphological refuse groups after low-temperature pyrolysis in CO2 (300–500 °C and 60 min) into CSF. The model considers cardboard, fabric, kitchen waste, paper, plastic, rubber, PAP/AL/PE (paper/aluminum/polyethylene) composite packaging pack, and wood, pyrolysis temperature, and residence time. The determination coefficients (R2) and Akaike information criteria were used for selecting the best model among four mathematical functions: (I) linear, (II) second-order polynomial, (III) factorial regression, and (IV) quadratic regression. For each RDF waste component, among these four models, the one best fitted to the experimental data was chosen; then, these models were integrated into the general model that predicts the HHV of CSF from the blends of RDF. The general model was validated experimentally by the application to the RDF blends. The validation revealed that the model explains 70–75% CSF HHV data variability. The results show that the optimal pyrolysis conditions depend on the most abundant waste in the waste mixture. High-quality CSF can be obtained from wastes such as paper, carton, plastic, and rubber when processed at relatively low temperatures (300 °C), whereas wastes such as fabrics and wood require higher temperatures (500 °C). The developed model showed that it is possible to achieve the CSF with the highest HHV value by optimizing the pyrolysis of RDF with the process temperature, residence time, and feedstock blends pretreatment.

show abstract

“…Apart from the analytical approach to the determination of HHV based on the proximate analysis, the characterization of the biomass in terms of its HHV using artificial neural networks is an active field of research (Olatunji et al 2020), (Abdulsalam et al 2020), (Ghugare et al 2014)... A nonlinear correlation between the higher heating value and the proximate, ultimate analysis has been proven (Aydinli et al 2017), (Dashti et al 2019). The ANNs trained on a relatively small set of data sets (25, referring to rice husks) outperformed empirical equations when compared to experimental HHV data (Yu et al 2014).…”

Section: Introductionmentioning

confidence: 99%

Comparing artificial neural network algorithms for prediction of higher heating value for different types of biomass

Jakšić

Jakšić²,

Guha³

et al. 2022

Soft Comput

View full text Add to dashboard Cite

The paper presents the comparative qualitative and quantitative analysis of twelve algorithms for training artificial neural networks (ANN) which predict the higher heating value (HHV) of biomass based on the proximate analysis (fixed carbon, volatile matter, and ash percentage). The twelve networks, with the same structure but different training algorithm, were fed with 318 experimental data triplets from literature for different biomass species and trained with 318 corresponding HHV values used for the supervised learning. Our comparative analysis showed that several algorithms resulted in ANNs generating outputs well correlated with the true measured values of the biomass HHV. Of those, Levenberg-Marquardt algorithm gives the best results in terms of mean squared error calculated on the training set data, while Bayesian regularization gives the best results in terms of regression. When applied to new datasets, unknown to the ANNs trained here, the highest accuracy of the HHV prediction was obtained by Conjugate Gradient, Powell/Beale Restarts training function. It ensured prediction based on the unknown datasets better than Levenberg-Marquardt algorithm. The described approach can be used for predicting the calorific values of different biomass species, including newly proposed ones, as well as for optimizing the HHV for both pure biomass and biomass mixtures.

show abstract

Application of Artificial Intelligence in the Prediction of Thermal Properties of Biomass

Cited by 6 publications

References 129 publications

Artificial neural networks in valorization process modeling of lignocellulosic biomass

Artificial neural networks in valorization process modeling of lignocellulosic biomass

The Prediction of Calorific Value of Carbonized Solid Fuel Produced from Refuse-Derived Fuel in the Low-Temperature Pyrolysis in CO2

Comparing artificial neural network algorithms for prediction of higher heating value for different types of biomass

Contact Info

Product

Resources

About