Fuel Cell Current Ripple Minimization in Real Time using PWM DC/DC Chopper based on ANN

The main aim of the present paper is the application of the selective harmonics elimination (SHE) in real time based on artificial neural network ANN with the DC/DC chopper interface for fuel Cell (FC) source to overcome the drawback of current ripple. This goal can be achieved in three steps, the off-line optimization of an objective function over a varying range of the average value of the output voltage of the DC/DC with the reduction or elimination of the selected low harmonics frequency, this task is achieved by using the optimization method of particle swarm optimization (PSO), then a stage of learning and training of the ANN based on back propagation is performed, then the obtained model is used in feed forward to ensure the generation of the switching patterns to fulfill the requirement of selected low frequency harmonics to be eliminated and to control the desired average value of the output voltage. Simulation results are presented for the validation of the proposed method.     «

The main aim of the present paper is the application of the selective harmonics elimination (SHE) in real time based on artificial neural network ANN with the DC/DC chopper interface for fuel Cell (FC) source to overcome the drawback of current ripple. This goal can be achieved in three steps, the off-line optimization of an objective function over a varying range of the average value of the output voltage of the DC/DC with the reduction or elimination of the selected low harmonics frequency,...     »