Distributed privacy‐preserving algorithm for microgrid energy management under time‐varying communication delays

Abstract This paper presents a consensus‐based distributed optimization algorithm to address the privacy‐preserving problem of microgrid energy management (MEM). The goal of MEM is to maximize the economic benefits of each device by adjusting the power generation/consumption while satisfying local physical constraints and supply–demand balance. A more general network model is considered, where each communication link has not only time‐varying, arbitrary, and heterogeneous communication delays, but also potential eavesdroppers attempting to invert privacy information based on the network information. To accommodate the above challenges, we design a fast convergence distributed algorithm and prove through matrix convergence theory that it can converge to a neighborhood of the optimal solution under bounded delay and noise. The size of the neighborhood depends on the upper bound of delay and the number of devices. By virtual node technique, we transform the original delay optimization problem into an equivalent delay‐free optimization problem in order to simplify the convergence analysis. In addition, we prove through polynomial analysis that the eavesdropper can not obtain any private information even if it obtains all the network information and algorithm structure, which implies the realization of privacy protection. Finally, we validate the proposed algorithm can converge under plug‐and‐play and large‐scale networks, and compare the performance of the algorithm under different maximum delays.