Privacy-preserving distributed secondary voltage control with predefined-time convergence for microgrids

Distributed secondary control is widely used in the hierarchical control structure of islanded microgrids. However, the information exchanged between distributed generators (DGs) may be intercepted by eavesdroppers, leading to the risk of privacy leakage and even data poisoning attacks that affect the stability of microgrids. Existing privacy-preserving distributed secondary control strategies suffer from low accuracy, high communication and computational overhead, and the poor convergence properties. Moreover, the objectives of multi-bus voltage regulation and proportional reactive power sharing cannot be achieved. To overcome these shortcomings, a privacy-preserving distributed average estimator is innovatively designed, where the states containing privacy information are decomposed into two parts based on the state decomposition method. The average estimation relies on the partial information exchange between neighboring DGs, thus avoiding the leakage of sensitive information. Furthermore, to address the difficulty that the state difference generated by the state decomposition leads to a long convergence time, a time-based generator is designed to effectively resolve the conflict between the privacy-preserving level and the convergence time. On this basis, a privacy-preserving distributed secondary voltage control is proposed, which preserves the privacy of the microgrid states while achieving average voltage regulation and proportional reactive power sharing within a predefined time, maintaining the system voltage stability and preventing the DGs from being heavily or lightly loaded. Finally, a hardware-in-the-loop platform for an islanded microgrid is built and the convergence and privacy-preserving performance of the proposed control strategy is verified.