The feedforward ANN controller is applied to the 1.26 kW rated fuel stack to achieve the peak power from the overall system at various operating temperature conditions of the fuel stack. In this controller, the backpropagation methodology is utilized to activate the neural network controller.
In this article, the electricity demand is illustrated in detail how it is increasing in an ascending manner. Conventional power systems are not useful to consumer demand. So, the fuel stack and PV sources are combined with the present available conventional sources to increase the availability of power to the consumers under different environmental conditions.
Conventional neural networks may not be suitable for hybrid PV/battery networks because of their lower working efficiency under various atmospheric temperature conditions. The radial basis functional neural network is utilized in this power supply network for the improvement of the convergence speed of the MPPT controller. The merits of this controller are more efficiency and less training data.
Most of the fuel stack works long lifetime duration when equated with the battery, and it works sufficiently with high operating efficiency. However, these systems do not give the linear response for peak load industrial applications. So, the ANFIS methodology is utilized in this network for running the operating point of the fuel stack near the actual MPP. The algorithm is monitored by applying the modified fluid search algorithm.
In this work, the neural network concept is utilized for the optimization of fuzzy membership rules. The proportional controller is interfaced in the fuzzy network for modifying the MPP location of the fuel stack system. The integrator in the controller helps to remove the dynamic oscillations of the overall system.
