Dynamic Stabilization of DC Microgrids with Predictive Control of Point-of-Load Converters

This paper investigates the possibility of deploying a finite control set model predictive control (FCS-MPC) algorithm for dynamic stabilization of a dc microgrid (MG) that supplies tightly regulated point-of-load (POL) converters. Within their control bandwidth, such converters behave as constant power loads (CPLs), where the MG sees them as impedances with a negative incremental resistance. Due to this characteristic, POL converters have a destabilizing impact that may cause large voltage oscillations or even a blackout of the whole MG. This paper proposes an active damping method realized by introducing a stabilization term in the cost function of the FCS-MPC algorithm that is used for regulation of the POL converter. This approach, on one hand, stabilizes a dc MG without implementing any additional active or passive components; thus, providing higher energy efficiency and better cost-effectiveness than methods that rely on such components. On the other hand, when compared to other approaches that focus on dc link stabilization via POL converter control, the proposed method has a significantly lower influence on the load voltage regulation performance. These findings are confirmed through comprehensive analytical investigation that shows how the proposed stabilization term affects the input impedance of the POL converter and the load voltage tracking performance. This is followed by experimental validation, where an FCS-MPC regulated uninterruptible power system inverter was used as a particular CPL example.