Secure Prespecified-Time Distributed Energy Management via Dual-Layer Privacy-Preserving Mechanism

This paper focuses on solving the energy management (EM) problem for microgrids in a fast and secure manner. To rapidly obtain the optimal power output combination, a flexible distributed EM algorithm over digraphs is proposed based on a novel prespecified-time convergence lemma, which enables the presetting of convergence time independent of initial states or algorithmic parameters. Considering the privacy protection of incremental costs (ICs), a new dual-layer privacy-preserving mechanism is devised to defend against internal honest-but-curious (HBC) distributed generators and external eavesdroppers alike. The first privacy-preserving layer is state decomposition, which operates throughout the entire scheduling horizon. The second privacy-preserving layer is edge-based additive perturbation, which operates during the coordinated scrambling phase. The state decomposition mechanism divides the ICs into two auxiliary variables, and the additive perturbation method can further protect the privacy of these auxiliary variables. Simulation results are provided to demonstrate the effectiveness of the proposed EM strategy.