Constraint-Aware Predictive Energy Coordination of Multiple Energy-Storage Converters for DC Bus Stabilization

Direct current (DC) bus voltage stability in grid-connected microgrids with multiple energy-storage converter branches is governed by the coupled dynamics of photovoltaic generation, heterogeneous battery storage, and the grid-connected inverter. Conventional independent and droop control methods allocate branch power mainly from local voltage or state-of-charge information, and therefore cannot fully account for converter limits, battery condition, or inverter current admissibility. This paper proposes a constraint-aware predictive energy coordination (CAPEC) method for DC bus stabilization. A DC bus energy model with variable losses is combined with one-step disturbance prediction and a safety governor that allocates storage power according to state of charge (SOC), state of health (SOH), power reserve, thermal margin, current capability, and ramp limits. The grid-connected inverter is treated as a constrained energy regulation path with active power saturation and antiwindup correction. Small-signal and energy function analyses show that CAPEC increases equivalent DC bus damping and yields bounded voltage recovery for feasible references and bounded prediction errors. MATLAB R2023b studies for combined photovoltaic and load variation, grid interface disturbance, and heterogeneous storage conditions show that CAPEC limits the maximum voltage deviation to 2.86 to 6.35 V. A reduced power laboratory experiment further verifies the Case B breaker disturbance and confirms stable DC bus behavior during unplanned grid interface switching.