Brain-Like Distributed Two-Layer SoC Balancing Control of Cyber-Physical Microgrid Clusters

As a typical cyber-physical system (CPS), AC microgrid (MG) clusters have attracted widespread attention due to the advantages such as high reliability, flexible scalability, and collaborative mutual support. A two-layer distributed SoC balancing control strategy is proposed for cyber-physical MG clusters containing multiple battery energy storage systems (BESSs). The strategy based on brain-like intelligence (BLI), aims to regulate the output voltage and frequency of all BESSs in each MG to their reference values, while also enabling active power sharing and state of charge (SoC) level balancing. The BLI controller, which is model-free, can swiftly learn and manage the complexities, nonlinearities, and uncertainties inherent in the MG model. The asymptotic convergence condition for the internal stability of the BLI controller has been established. Furthermore, a sparse two-layer communication network is created by connecting one or more BESSs from each MG in the lower network to form an upper network, and the secondary and tertiary response matching conditions for the closed-loop MG clusters are derived. Finally, the effectiveness and robustness of the proposed strategy is validated through multiple real-time simulation cases of the modified IEEE 34-bus system by using OPAL-RT.