Power-quality-enhanced control scheme applied in V2G converters: Unified gains optimization for previewable reference feedforward, repetitive learning, and state feedback

The rapid proliferation of electric vehicles (EVs) faces challenges due to diversified battery voltage levels and power quality degradation in vehicle-to-grid (V2G) chargers under distorted grid conditions. This paper proposes a hybrid hierarchical control scheme for two-stage V2G converters, offering two primary advantages: the front-stage dual-active-bridge (DAB) converter could stabilize the DC-link voltage and accommodate wide-ranging battery voltages; meanwhile, a unified linear matrix inequality (LMI) framework is established for the rear-stage DC–AC converter to systematically synthesize all control gains, which integrates a resonant-suppressing state feedback controller to reduce parameter sensitivity, repetitive learning to attenuate multi-order harmonics in the charger current, and previewable reference current feedforward to enhance tracking precision. Experimental results demonstrate superior performance over conventional methods: total harmonic distortion is reduced by 55% (from 8.72% to 3.97% in V2L mode), zero steady-state error is achieved under distorted grid voltages (V2G mode), and tracking accuracy improves by over 86% under exceeding 30% parameter deviations (V2L/V2G mode), alongside enhanced grid harmonic suppression.