基于变流器输出阻抗的直流微电网下垂并联系统振荡机理与稳定边界分析

DC microgrids are of great significance for fully using renewable energy and actively responding to the carbon peaking and carbon neutrality goals. In DC microgrids, power electronics converters are usually connected to the electrical power generation units and the DC bus. These converters interact with each other through parallel connections, which increases the instability risk of the system. However, existing research regarding the system stability of DC microgrids only focuses on the influence of the interaction between the source subsystem and the load subsystem. It does not consider the difference among parallel converters in the source subsystem and thus ignores the influence of the difference among the source converters on the system stability. The difference among the source converters will cause the difference in their output impedances, which will introduce the system oscillation. Therefore, this paper focuses on the influence of the interaction among the Buck converters on the system stability, analyzes the oscillation mechanism of the parallel converters, and proposes a solution for the stability boundary. Firstly, the small-signal model of the output impedance of the single Buck converter under droop control is established in this paper, and the accuracy of the output impedance is verified by the frequency sweep in SABER. Secondly, the controller, main circuit, and parasitic parameters are analyzed to the output impedance model established, revealing the oscillation mechanism of the droop-controlled Buck converters in parallel. Thirdly, through qualitative analysis, the output impedance is further simplified. Fourthly, the stability boundary regarding the key parameters of the parallel converters is calculated by combining the oscillation mechanism revealed and the stability criterion. Finally, the experimental platform of parallel Buck converters is built to verify the correctness and effectiveness of the theoretical analysis. The experimental results of the parallel converters show that when the proportional coefficient k_pi2 is less than 0.09, the time domain waveform of the parallel converters is stable. Then increase the proportional coefficient k_pi2 to 0.09 and 0.1, the output voltage oscillates, and the parallel system becomes unstable. The oscillation period is 1.3 ms, corresponding to the theoretical analysis of 781 Hz. Due to the slight difference between the parameters of the experimental platform and the theoretical analysis, the theoretical value of k_pi2 is also slightly different from the experimental value when it is critically stable. However, the error is within an acceptable range, and the oscillation frequency is consistent with the theoretical analysis, which verifies the effectiveness and correctness of the proposed oscillation mechanism and stability boundary. The following conclusions can be drawn. This paper takes the parallel droop-controlled Buck converters as the research object, and the influence of the difference among Buck converters on the system stability is investigated. The output impedance of the Buck converter is characterized as follows. The amplitude-frequency curve has a resonance peak, and a phase mutation appears at the resonance peak frequency. The resonance peak frequency is mainly affected by key parameters such as the filter inductance L_f, the filter capacitor C_f and the current controller proportional coefficient k_pi. The mechanism of system oscillation is revealed when the key parameters, including filter inductances L_f and current controller parameters k_pi, of the converters are inconsistent and cause the phase difference between the output impedance of converters at the intersection frequency of amplitude-frequency curves exceeds 180°. Finally, a segmented analytical formula of the output impedance and the stability boundary of the key parameters of the system is proposed. The experiments verify the correctness of the theoretical analysis.