Metaheuristic-based Near-Optimal Fractional Order PI Controller for On-grid Fuel Cell Dynamic Performance Enhancement

Abstract

Fractional-order PI (FOPI) is considered a possible efficient alternative controller to classical PI controller for fuel cell (FC) on-grid systems applications. This paper aims to enhance the dynamic performance of grid-connected FC systems while employing different metaheuristic optimization techniques to FOPI controllers and consequently improve the power quality and stability of the system. The proposed metaheuristic techniques are modified flower pollination algorithm (MFPA), harmony search (HS), and electromagnetic field optimization (EFO). In the study, two FOPI-based regulators are suggested for controlling the current and the power flow between the FC and the network. For this purpose, the inverter's pulse width modulation (PWM) regulation is performed while integrating FC systems into the network. The three metaheuristic optimizers are used for the optimal gain scheduling of the FOPI controller parameters while minimizing the absolute value of the error as the cost function. To demonstrate how efficient the FOPI controllers are, fair comparisons for the on-grid FC-based system performance while proposing the different optimized advanced FOPI and conventional PI controllers are analyzed. According to the simulation results, the most efficient metaheuristic technique is identified based on an impartial defined criterion regarding the dynamic behaviors’ steady-state error, overshoot besides both setting and execution time. Through the suggested criterion, the EFO is considered the best optimization approach. Through the numerical analysis, the power quality assessment of the optimally controlled overall FC on-grid systems is illustrated through voltage sag/swell studies.