Artificial neural network models for biomass gasification in fluidized bed gasifiers

Puig-Arnavat, Maria; Hernández, J.A.; Bruno, Joan Carles; Coronas, Alberto

doi:10.1016/j.biombioe.2012.12.012

Cited by 147 publications

(85 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Two different types of ANN based data-driven models have been developed for the prediction of gas production rate and heating value of gas in coal gasifiers (Chavan et al, 2012). Recently, ANN based predictive tools have been used in fluidized bed gasifiers to predict the syngas composition and gas yield (Puig-Arnavat et al, 2013). The ANN technique has been applied in the gasification area and has shown better results compared to the conventional process modelling approaches.…”

Section: As Per the Eu's New Directive Eachmentioning

confidence: 99%

See 1 more Smart Citation

Artificial neural network based modelling approach for municipal solid waste gasification in a fluidized bed reactor

et al. 2016

View full text Add to dashboard Cite

In this paper, multi-layer feed forward neural networks are used to predict the lower heating value of gas (LHV), lower heating value of gasification products including tars and entrained char (LHVp) and syngas yield during gasification of municipal solid waste (MSW) during gasification in a fluidized bed reactor. These artificial neural networks (ANNs) with different architectures are trained using the Levenberg-Marquardt (LM) back-propagation algorithm and a cross validation is also performed to ensure that the results generalise to other unseen datasets. A rigorous study is carried out on optimally choosing the number of hidden layers, number of neurons in the hidden layer and activation function in a network using multiple Monte Carlo runs. Nine input and three output parameters are used to train and test various neural network architectures in both multiple output and single output prediction paradigms using the available experimental datasets. The model selection procedure is carried out to ascertain the best network architecture in terms of predictive accuracy. The simulation results show that the ANN based methodology is a viable alternative which can be used to predict the performance of a fluidized bed gasifier.2

show abstract

Section: As Per the Eu's New Directive Eachmentioning

confidence: 99%

“…A brief overview of different modelling approaches and their pros and cons is presented in Table 1. or any other process parameter which are deemed necessary (Puig-Arnavat et al, 2013).…”

Section: As Per the Eu's New Directive Eachmentioning

confidence: 99%

Artificial neural network based modelling approach for municipal solid waste gasification in a fluidized bed reactor

et al. 2016

View full text Add to dashboard Cite

show abstract

“…temperature, pH, enzyme and substrate loadings and a combination of these parameters) that provided the most valuable information for apple pomace hydrolysis were selected for the development of the ANN model. The selection of the most appropriate parameters for ANN modelling is considered of paramount importance for prediction of the hydrolysis process (Puig-Arnavat et al 2013;Rivera et al 2010). The constructed ANN was assessed for its accuracy for generalisation and predictive ability using R 2 value and MSE values for the outputs.…”

Section: Annmentioning

confidence: 99%

Using an artificial neural network to predict the optimal conditions for enzymatic hydrolysis of apple pomace

et al. 2017

View full text Add to dashboard Cite

The enzymatic degradation of lignocellulosic biomass such as apple pomace is a complex process influenced by a number of hydrolysis conditions. Predicting optimal conditions, including enzyme and substrate concentration, temperature and pH can improve conversion efficiency. In this study, the production of sugar monomers from apple pomace using commercial enzyme preparations, Celluclast 1.5L, Viscozyme L and Novozyme 188 was investigated. A limited number of experiments were carried out and then analysed using an artificial neural network (ANN) to model the enzymatic hydrolysis process. The ANN was used to simulate the enzymatic hydrolysis process for a range of input variables and the optimal conditions were successfully selected as was indicated by the R 2 value of 0.99 and a small MSE value. The inputs for the ANN were substrate loading, enzyme loading, temperature, initial pH and a combination of these parameters, while release profiles of glucose and reducing sugars were the outputs. Enzyme loadings of 0.5 and 0.2 mg/g substrate and a substrate loading of 30% were optimal for glucose and reducing sugar release from apple pomace, respectively, resulting in concentrations of 6.5 g/L glucose and 28.9 g/L reducing sugars. Apple pomace hydrolysis can be successfully carried out based on the predicted optimal conditions from the ANN.

show abstract

“…A number of studies had been carried out in the past on modelling of the thermochemical conversation of various biomass materials (Muilenburg et al 2011;Bates et al 2016). Phenomenological models that represent the various chemical processes in each stage of gasification using mathematical relations (Muilenburg et al 2011;Sulaiman et al 2012;Bates et al 2016) are more commonly applied than the black box modelling approach that use Artificial Neural Network (ANN) (Puig-Arnavat et al 2013;Rodrigues et al 2016). These models can be placed into two main classes: equilibrium models that assume all major reactions of pyrolysis and gasification to reach chemical equilibrium during the gasification process, and kinetic models, which involve kinetic rate equations.…”

Section: Introductionmentioning

confidence: 99%

Modeling and Parametric Study for Maximizing Heating Value of Gasification Syngas

et al. 2017

View full text Add to dashboard Cite

There are a number of experimental and theoretical studies on the energy conversion of oil palm derivative biomass. Moreover, the potential of this abundant biomass residue for renewable energy in major producing countries in Southeast Asia has been well documented. In this study, the results of an equilibrium model of downdraft gasification of oil palm fronds (OPF), developed using the Aspen Plus chemical process simulator software, and its validation are presented. In addition, an optimization of the major output parameter of importance (i.e., the higher heating value of syngas) with respect to the main operating parameters (i.e., temperature, equivalence ratio (ER), and moisture content) was performed. The response surface method (RSM) was used to determine the mathematical relationship between the response of interest, which was the heating value of syngas, and the operating conditions. This method was used to further determine the conditions that would lead to optimum higher heating values of syngas. Optimum values identified by RSM were: oxidation zone temperature of 1000 °C, moisture content in the range of 4%, and an equivalence ratio of 0.35. These optimum operating conditions and the corresponding higher heating value of syngas were found to correspond with the experimental results.

show abstract

Artificial neural network models for biomass gasification in fluidized bed gasifiers

Cited by 147 publications

References 14 publications

Artificial neural network based modelling approach for municipal solid waste gasification in a fluidized bed reactor

Artificial neural network based modelling approach for municipal solid waste gasification in a fluidized bed reactor

Using an artificial neural network to predict the optimal conditions for enzymatic hydrolysis of apple pomace

Modeling and Parametric Study for Maximizing Heating Value of Gasification Syngas

Contact Info

Product

Resources

About