Dynamic double-frequency power minimization in multi-converter hybrid DC–AC buildings with single-phase inverters

This paper introduces an enhanced control strategy to mitigate DC bus double-frequency power oscillations in a building-integrated DC–AC hybrid microgrid composed of renewable energy sources, energy storage systems, and AC building dwelling loads. Integrating multiple single-phase inverters in a microgrid (MG) produces a double-frequency power oscillation at the DC side, adversely affecting power quality and system stability. To address this issue, an enhanced control strategy is proposed to optimize the voltage reference phase angles for each AC single-phase load inverter by considering factors like load power factor and apparent power. The proposed strategy uses established optimization techniques (i.e., Gradient Descent, Nelder–Mead) to dynamically adjust the voltage phase angles. A total of 10 techniques are evaluated for comparison to determine the most appropriate optimization technique for an increased number of connected inverters. The best technique resulting from the benchmarking is evaluated experimentally. The proposed control strategy increases the DC bus capacitor’s lifetime, reduces switching stress, and improves MG stability by mitigating the double-frequency power oscillation. The feasibility of the proposed method is validated through experimental implementation, considering a real-time lead controller, communications, and real load profiles, including hardware-in-the-loop and respective controllers. The proposed control achieves up to a 93% reduction in low-frequency DC bus power oscillations and also extends the capacitor lifetime 2.5 times. • A novel optimization-based control strategy that mitigates the double-frequency DC power oscillations by dynamically adjusting the voltage reference phase angle of the single-phase inverters. The method is scalable to over 30 inverters without increasing computational burden or complexity, differently than the limited number of inverters from previous strategies. • Scalable solution for independent inverter control and simple expansion. • Comparative study of 10 optimization algorithms for systems up to 30 inverters. • Definition of the best algorithm, highlighting control robustness and scalability. • Real-time experimental validation including switching effects and control dynamics. • Centralized implementation including real communication delays and demand profiles. • Benefits include longer capacitor life, lower switching stress, modularity, and improved MG stability.