A decentralized/distributed cluster-based predefined-time control for DC networked microgrids in the presence of constant power loads

Addressing the challenges introduced by the growing scale of Networked MicroGrids (NMGs) and the rising penetration of inherently destabilizing Constant Power Loads (CPLs), this paper designs a novel decentralized/distributed cluster-based control framework that combines both advantages of decentralized control and distributed control. Different from the existing fully decentralized or fully distributed control architectures, the proposed control framework performs a leader–follower structure within each NMG while maintaining a decentralized scheme among different NMGs. To guarantee predefined-time control performance and avoid the need for step-by-step stabilization, a decentralized/distributed cluster-based predefined-time controller is designed to simultaneously achieve current sharing and voltage restoration by introducing an extra gain in the coupling current and voltage elements. The stability analysis on the specific multi-zero-root Laplacian matrix introduced by the proposed control framework is given and a more relaxed upper bound on the allowed CPLs is also derived with respect to existing methods. Both theoretical proof and simulation results show that the proposed control law can not only allow a large proportion of CPLs to connect in DC NMGs, but also achieves a fast predefined-time control performance.