Tertiary Control Based on Non-Parametric Model Prediction for DC Microgrid Cluster

Multiple geographically adjacent DC microgrids can be interconnected to form a network system to improve the economics, inherent, and reliability of the power supply. To manage such a complex network, the hierarchical control paradigm can be adopted, where primary and secondary controllers deal with voltage and current regulation within a single DCMG whereas tertiary controller focuses on power exchange among the interconnected DCMGs in the cluster. Existing tertiary controllers are generally implemented with conventional PI, which is difficult to adapt because of the over-burdened task of modeling a complex network such as DCMG. To address this issue, this paper proposes a tertiary control strategy based on non-parametric model prediction for DCMGs. The performance of the proposed control strategy is verified by simulations and experimental results.