An Accurate Reactive Power Sharing Strategy Based on Adaptive Virtual Impedance Optimization

For isolated microgrids, the key task is to achieve precise power-sharing according to the rated capacity ratio of the distributed generators (DGs). However, due to the mismatch between feeder impedances, conventional droop control strategies become ineffective to achieve precise reactive power sharing. Such drawback may put the entire microgrid at risk, as it will cause large loop current between inverters, and even leads to current overload of certain inverters. In this paper, we mathematically analyze the cause of inaccurate reactive power sharing when applying droop control techniques and then, we propose a low communication-dependent strategy based on adaptive virtual impedance optimization to achieve accurate reactive power sharing. First of all, the actual feeder impedances are identified online by using the recursive least squares algorithm with forgetting factor, which is passive and ensures high-quality power supply to the public loads. Second, the proposed strategy uses the identification results of feeder impedances to realize the optimal adjustment of virtual complex impedance for each inverter. This strategy prevents additional deteriorations of the microgrid voltages in comparison to existing adaptive virtual impedance strategies. Additionally, the proposed control is credible and robust to communication faults. Finally, the feasibility and effectiveness of the proposed strategy is verified in both normal and communication fault conditions by simulation results.