Power Management with Event-Triggered Predictive Function Control for DC Microgrid Clusters

A DC microgrid cluster (DCMGC) is a networked power grid that consists of multiple small-scale DC microgrids (DCMGs), and every single DCMG (sub-grid) can be composed of multiple heterogeneous units, e.g., distributed energy resources like photovoltaic (PV) and wind turbine (WT), energy storage systems (ESS), and local loads. The DCMGC is such a complex network with multiple control objectives that requires control at both system-level and sub-grid level, which suits well with model predictive control (MPC) algorithms. However, the MPC suffers from the high computation burden, especially dealing with complex control plant models like DCMGCs. In this paper, an event-triggered predictive function control (ET-PFC) is proposed for power management of DCMGCs, maintaining excellent control performance of PFC algorithm while saving computing resources in the control process. Further, a universal low-cost microcontroller (STM32) solution is presented for implementation of the ET-PFC scheme and a low-voltage DCMGC laboratory prototype that consists of three DCMGs (six converters) is built to validate the controller implementation by hardware experiment results.