The model predictive control (MPC) is a well-accepted method for controlling power electronic converters. This paper presents a tailored MPC approach in which the internal and external dynamics of the dc/dc modular multilevel converter (MMC) are integrated into the MPC algorithm. The proposed MPC approach introduces three control objectives to have full control over the internal and external dynamics. Each of the designed cost functions includes one primary term regulating one of the control objectives and one secondary term improving the converter performance. Unlike the conventional control approach based on multiple proportional-integral (PI) controllers, the proposed approach provides a straightforward way to design the control parameters. The operation of the presented MPC approach is thoroughly investigated and compared to that of the PI-based controller. Comparative simulation studies confirmed that the proposed MPC approach reduces the ac circulating current in the steady-state operation compared to the conventional PI-based control. In the transient mode, the MPC approach offers much smoother and faster responses to the changes in the power reference. The performance of the dc/dc MMC controlled by the proposed MPC approach under parametric uncertainty is investigated, and improved performance is obtained compared to the conventional PI-based control.

中文翻译：

---

非隔离DC/DC模块化多电平变换器的模型预测控制提高动态响应


模型预测控制（MPC）是一种广为接受的电力电子转换器控制方法。本文提出了一种定制的 MPC 方法，其中将 DC/DC 模块化多电平转换器 (MMC) 的内部和外部动态集成到 MPC 算法中。所提出的 MPC 方法引入了三个控制目标，以完全控制内部和外部动态。每个设计的成本函数都包括调节控制目标之一的一项主要项和提高转换器性能的一项次要项。与基于多个比例积分（PI）控制器的传统控制方法不同，所提出的方法提供了一种直接的方法来设计控制参数。我们对所提出的 MPC 方法的操作进行了彻底的研究，并与基于 PI 的控制器进行了比较。比较仿真研究证实，与传统的基于 PI 的控制相比，所提出的 MPC 方法减少了稳态运行中的交流循环电流。在瞬态模式下，MPC 方法对功率参考的变化提供更平滑、更快速的响应。研究了在参数不确定性下由所提出的 MPC 方法控制的 dc/dc MMC 的性能，与传统的基于 PI 的控制相比，获得了改进的性能。