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

Abstract

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

Comparison of the different artificial neural networks in prediction of biomass gasification products

“…ANNs are universal predictors that take advantage of previously obtained experimental data. So that it is preferred over other theoretical and empirical models whose primary goal is predictive accuracy . ANN is an architecture that contains many different layers including organized neurons, and through the weights, the neurons in each layers are connected to another layer.…”

“…Depending on the complexity of the gasification, these models may be 3D fluid dynamical, artificial neural networks, artificial intelligence, kinetic‐based, or less complex equilibrium models . Even though none of the mathematical model is perfect for expressing or representing the gasification system, there is still a need for developing mathematical models . For computational fluid dynamics (CFD) models, a set of energy, momentum, and mass equations are solved to find the distribution of various parameters such as concentration and temperature over a certain region of a gasifier .…”

“…An ANN model was performed for a fluidized bed gasifier with operating conditions of 400°C to 800°C and the ER in the range of 0.2 to 0.6 . In addition to prediction of syngas yield, Pandey et al selected the LHV of gasification products (entrained char and tars) as output variables in the ANN model for MSW gasification in a fluidized bed gasifier . The proposed method iteratively changes hyperparameters such as number of neurons on the hidden layers and activation functions on the network to choose best fitting model.…”

Evolutionary Prediction of Biohydrogen Production by Dark Fermentation

“…Furthermore, a quadratic model might not provide the proper response to describe the variables. Hence, computational intelligence methods have a significant role in this matter …”

Determining Mechanical and Physical Properties of Phospho-Gypsum and Perlite-Admixtured Plaster Using an Artificial Neural Network and Regression Models

“…Choosing the right ANN depends on the type of the application and data representation of a given problem [19,11,20].…”