Virtual inertia for suppressing voltage oscillations and stability mechanisms in DC microgrids

Abstract

Renewable energy sources (RES) are gradually penetrating power systems through power electronic converters (PECs), which greatly change the structure and operation characteristics of traditional power systems. The maturation of PECs has also laid a technical foundation for the development of DC microgrids (DC-MGs). The advantages of DC-MGs over AC systems make them an important access target for RES. Due to the multi-timescale characteristics and fast response of power electronics, the dynamic coupling of PEC control systems and the transient interaction between the PEC and the passive network are inevitable, which threatens the stable operation of DC-MGs. Therefore, this dissertation focuses on the study of stabilization control methods, the low-frequency oscillation (LFO) mechanism analysis of DC-MGs and the state-of-charge (SoC) imbalance problem of multi-parallel energy storage systems (ESS). Firstly, a virtual inertia and damping control (VIDC) strategy is proposed to enable bidirectional DC converters (BiCs) to damp voltage oscillations by using the energy stored in ESS to emulate inertia without modifications to system hardware. Both the inertia part and the damping part are modeled in the VIDC controller by analogy with DC machines. Simulation results verify that the proposed VIDC can improve the dynamic characteristics and stability in islanded DC-MG. Then, inertia droop control (IDC) strategies are proposed for BiC of ESS based on the comparison between conventional droop control and VIDC. A feedback analytical method is presented to comprehend stability mechanisms from multi-viewpoints and observe the interaction between variables intuitively. A hardware in the loop (HIL) experiment verifies that IDC can simplify the control structure of VIDC in the promise of ensuring similar control performances. Subsequently, a multi-timescale impedance model is established to clarify the control principle of VIDC and the LFO mechanisms of VIDC-controlled DC-MG. Control loops of different timescales are visualized as independent loop virtual impedances (LVIs) to form an impedance circuit. The instability factors are revealed and a dynamic stability enhancement method is proposed to compensate for the negative damping caused by VIDC and CPL. Experimental results have validated the LFO mechanism analysis and stability enhancement method. Finally, an inertia-emulation-based cooperative control strategy for multi-parallel ESS is proposed to address the SoC imbalance and voltage deviation problem in steady-state operation and the voltage stability problem. The contradiction between SoC balancing speed and maintaining system stability is solved by a redefined SoC-based droop resistance function. HIL experiments prove that the proposed control performs better dynamics and static characteristics without modifying the hardware and can balance the SoC in both charge and discharge modes.