Multi-Condition PSO-FOPI Current-Sharing Control for Parallel LLC Resonant Converters under Parameter Deviations

Abstract

Parallel LLC resonant converters are suitable for low-voltage and high-current applications. However, deviations in the resonant inductor, resonant capacitor, and magnetizing inductor cause gain differences among phases, resulting in output-current imbalance. To address the difficulty of achieving satisfactory multi-condition dynamic performance with a dual-loop controller under unknown resonant-parameter deviations, this paper proposes a multi-condition particle swarm optimization fractional-order proportional-integral (PSO-FOPI) current-sharing control method. The proposed method adopts a common-mode/differential-mode frequency dual-loop structure, in which common output-voltage regulation and interphase current-sharing regulation are mapped to the common-mode and differential-mode frequencies, respectively. Meanwhile, the worst-case time-domain performance over typical parameter-deviation conditions is used as the fitness function, and particle swarm optimization (PSO) is employed to offline tune the dual-loop FOPI controller parameters. Simulation results show that, under six typical parameter-deviation conditions and within a ±15% Monte Carlo random parameter-deviation range, the controller maintains favorable voltage and current-sharing dynamic performance. Hardware experiments further verify the practical operating feasibility of the proposed control method.