Distributed optimal power management for smart homes in microgrids with network and communication constraints

A distributed resilient event-triggered economic power dispatching problem (EPD) for multiple smart homes in microgrids considering network constraints is studied in this paper. To address the EPD, a privacy-aware distributed weighted optimization algorithm (DWOA) with a dynamic event-triggered communication mechanism that avoids Zeno behavior is proposed. Through Lagrangian optimality analysis, the EPD problem, initially constrained by network constraints, is transformed into an unconstrained problem by incorporating global coupling multipliers. Subsequently, a distributed calculation strategy of global coupling multipliers is designed. Moreover, a distributed resilient optimization strategy and communication strategy are designed to solve the EPD when communication links are threatened by denial of service (DoS) cyber-attacks. Furthermore, predictive information is utilized to adjust the eigenvalues of weighted control matrix aiming at enhancing the resilience to DoS attacks. To, therefore, investigate the convergence and optimality of the proposed optimization algorithms, four theorems are proposed, which show that the EPD can be solved when the network constraints as well as DoS attacks are considered. Simulation results show that the proposed DWOA solves the EPD with reduced communications for ideal links, even if network constraints are considered. For the case of DoS attacks, the proposed resilient DWOA can also solve the EPD as long as the attack frequency and duration are within the limits.