Event-Triggered Predictive Function Control of DC Microgrid Clusters

DC microgrid cluster (DCMGC) is a dynamic network of multiple DC microgrids (DCMGs) that interconnected through tie-lines. By effective power flow control among the interconnected multiple DCMGs, the overall DCMGC is capable of achieving the most in terms of reliability, efficiency, and cost reduction. But the high penetration of distributed energy resources and power electronics converters render the control design of the complex DCMGCs a challenge task. The model predictive control (MPC) has been proven as a promising solution in the control of DCMGCs. However, the mechanism of the receding horizon optimization of MPC impose much stress onto the online computation and leads to the waste of communication resources as well, especially in the steady-state operation of the systems. To address this issue, an event-triggering is added to the MPC algorithm in the power flow control of DCMGCs, and the control command is only updated subject to the triggering events. Hence, the high frequency constant online computation and optimization will be significantly reduced. A hardware-in-the-loop (HIL) prototype of the DCMGC that consists of three DCMGs is built to verify the feasibility of the proposed method.