Rigdelet Neural Networks-based Maximum Power Point Tracking for a PEMFC connected to the network with Interleaved Boost Converter optimized by Improved Satin Bowerbird Optimization

Abstract

This paper proposes a new control policy for optimal control of the 3-phase system of PEMFC connected to the grid. This also includes a 3-phase high step-up Interleaved Boost Converter (IBC) to amplify the PEMFC outputted voltage. To control the PEMFC system, Maximum Power Point Tracking (MPPT) based on Rigdelet Neural Networks (RNN) has been utilized, and to improve this controller, an improved version of the Satin Bowerbird Optimization (ISBO) algorithm has been utilized. The main advantage of the proposed improved version is modifying the convergence weakness and fixing convergence in chaos theory. The method is then validated by performing it one time on a standalone PEMFC system and another time on a grid-connected PEMFC system. Simulation results indicate that based o the IBC converter, better results with lower current ripples can be achieved. Also, the method has the ability to feed to both active and reactive powers by keeping stable the sudden temperatures. Final results have been also put in comparison with two different latest techniques to indicate the technique proficiency.

Keywords

• Polymer electrolyte membrane fuel cells
• MPPT
• IBC
• Control
• Rigdelet Neural Networks
• Grid-connected