模块化多电平换流器的降阶小信号模型研究

Research on Reduced-order Small-signal Model of Modular Multilevel Converter

建立模块化多电平换流器（modular multilevel converter,MMC）的小信号模型时,通常会考虑子模块电容电压波动、内部环流等状态变量,但却提高了小信号模型的阶数。为了减少多端MMC直流网络进行小信号分析时的模型阶数,该文首先建立MMC全阶小信号模型;然后基于平衡理论和Hankel奇异值降阶思想,提出MMC全阶小信号模型的一般降阶方法;建立MMC整流站和逆变站的有效降阶小信号模型;并以三端MMC-HVDC系统为例对降阶前后系统的动态响应特性和不稳定特征根轨迹进行了对比。仿真结果表明：通过选取适当的阶数,MMC降阶模型与全阶模型具有一致性;直流系统在应用降阶小信号模型后,系统的降阶模型与原系统的全阶小信号模型具有相同的动态特性,并且同样能够对系统不稳定进行预估。

To study the small-signal model of modular multilevel converter（MMC）, the inner state variables including sub-module capacitor voltage fluctuations, the inner loop current are usually considered. However, this will greatly increase the order of MMC. To decrease the order of the small-signal model of multi-terminal MMC-MTDC, this paper first derived the full-order small signal model of MMC, and then presented the general model-reduction method for MMC based on balanced realization and Hankel singular value theories. With the presented approach, two valid reduced-order small-signal models were derived for rectifier and inverter, respectively. Taking a three-terminal MMC-HVDC grid for example, the dynamic response and the root locus of the systems before and after model reduction were compared and analyzed. The simulation shows that by keeping appropriate order, the reduced-order model of MMC performs consistently with full-order model of MMC; with the proposed model-reduced method, MMC based HVDC grid has the same dynamic response with full-order and reduced-order system models, and the reduced-order model can correctly predict the instability of the system.