• Corpus ID: 2950854

Neural Network-Based Modeling of PEM fuel cell and C ontroller Synthesis of a stand-alone system for residential application

@inproceedings{Mammar2012NeuralNM,
  title={Neural Network-Based Modeling of PEM fuel cell and C ontroller Synthesis of a stand-alone system for residential application},
  author={Khaled Mammar and Abdelkader Chaker},
  year={2012}
}

• Khaled Mammar, A. Chaker
• Published 2012
• Engineering

The paper is focused especially on presenting possibilities of applying artificial neural networks at creating the optimal model PEM fuel cell. Various ANN approaches have been tested; the back-propagation feed-forward networks show satisfactory performance with regard to cell voltage prediction. The model is then used in a power system for residential application. This models include an ANN fuel cell stack model, reformer model and DC/AC inverter model. Furthermore a neural network (NNTC) and fuzzy logic (FLC) controllers are used to control active power of PEM fuel cell system. The controllers modifies the hydrogen flow feedback from the terminal load. The validity of the controller is verified when the fuel cell system model is used in conjunction with the NNT controller to predict the response of the active power to: (a) computer-simulated step changes in the load active and reactive power demand, and (b) actual active and reactive load demand of a single family residence. Simulation results confirmed the high performance capability of the neural network (NNTC) to control power generation.