The deep feedforward artificial neural network (DFF ANN) is proposed in this study to model the new Molten Borate Fuel Cell behavior. This modeling approach is well-known for being a potent tool for dealing with complicated modeling and prediction problems. This study presented many network designs for various operation circumstances. In most cases, the ANN was utilized to model and optimize the SOFC, MCFC, and PEMFC, but the design and performance of the MBFC were not investigated. The described innovative type of fuel cells - molten borate fuel cells - are designed and optimized using a deep learning algorithm. The most advanced model can provide a dynamic forecast of fuel cell operation while taking thermal-flow and electrolyte material factors into account. With an average inaccuracy of 0.3 percent, all of the models indicated an accurate operation. Furthermore, an ANN-based technique may be used to improve the cell operating parameters. Additionally, the operating composition of a novel molten borate electrolyte might be improved in terms of electrochemical performance of the fuel cell.
Keywords
- Molten Borate Fuel Cell
- Hydrogen
- Artificial neural networks
- Mathematical modelling
- Optimization
Acknowledgment
Studies were funded by ENERGYTECH-1 project granted by Warsaw University of Technology under the program Excellence Initiative: Research University (ID-UB).