Search

Summary: This article shows the teaching processes of artificial neural networks that are used to model the molten carbonate fuel cell (MCFC). Researchers model MCFCs to address a variety of issues across a range of complexities, from simply gauging the effect of temperature through to a complete model with 14 input parameters. The architecture of the model is a triple layer network with one hidden layer containing three neurons. The activation function used for the hidden layer was a hyperbolic tangent, with the last layer being based on linear function. We produced various network configurations, mostly networks containing one hidden layer. Models map the work of a real fuel cell with an average error in the range of 2.4% to 4.6%. The model we created guided the optimization of the thermal‐flow and construction parameters of the MCFC. Commercially available software was used to build the model and optimize the operating parameters. The selected objective functions were the efficiency of electricity production and the power density obtained from the cell's surface. The results obtained serve as pointers for possible changes in fuel cell operation and could lead to some structural changes being made. [ABSTRACT FROM AUTHOR]

The idea of control strategy of SOFC operating to meet demand of a public utility building was presented. The strategy was formulated with the support of Artificial Neural Network. The network was used to predict the demand for electricity. The calculations were carried out on the example of a building of the Institute of Heat Engineering Warsaw University of Technology. The control strategy is influenced by various factors depending on changes in market conditions and operating characteristics of the cell. We can define different objective functions eg: working for own needs, for maximum profit and maximum service life. The article presents a simulation of SOFC operation for demand profile of the IHE building from the selected time period. [ABSTRACT FROM AUTHOR]

There are many mathematical models of the singular solid oxide fuel cell (SOFC). SOFC performance modelling is related to the multi-physic processes taking place on the fuel cell surfaces. Heat transfer together with electrochemical reactions, mass and charge transport are conducted inside the cell. There are many parameters which impact the cell working conditions, e.g. electrolyte material, electrolyte thickness, cell temperature, inlet and outlet gas compositions at anode and cathode, anode and cathode porosities etc. The Artificial Neural Network (ANN) can be applied to simulate an object's behaviour without an algorithmic solution merely by utilizing available experimental data. The ANN is used for modelling singular cell behaviour. The optimal network architecture is shown and commented. The error back-propagation algorithm was used for an ANN training procedure. [ABSTRACT FROM AUTHOR]

Abstract: The Artificial Neural Network (ANN) can be applied to simulate an object''s behaviour without an algorithmic solution merely by utilizing available experimental data. The ANN is used for modelling singular cell behaviour. The optimal network architecture is shown and commented. The error backpropagation algorithm was used for an ANN training procedure. The ANN based SOFC model has the following input parameters: current density, temperature, fuel volume flow density (mlmin−1 cm−2), and oxidant volume flow density. Based on these input parameters, cell voltage is predicted by the model. Obtained results show that the ANN can be successfully used for modelling the singular solid oxide fuel cell. The self-learning process of the ANN provides an opportunity to adapt the model to new situations (e.g. certain types of impurities at inlet streams etc.). Based on the results from this study it can be concluded that, by using the ANN, an SOFC can be modelled with relatively high accuracy. In contrast to traditional models, the ANN is able to predict cell voltage without knowledge of numerous physical, chemical, and electrochemical factors. [Copyright &y& Elsevier]