Predictive Duty Cycle Control for Four-Leg Inverters With ${LC}$ Output Filter

Abstract

Three-phase four-leg inverter is a well-known solution to handle unbalanced and nonlinear loading conditions in three-phase transformerless uninterruptible power supply applications. For such power converters, finite control set model predictive control (FCS-MPC) is considered as a promising control alternative due to its simplicity, fast dynamic response, and capability to include nonlinearities and constraints. However, the absence of modulator in FCS-MPC implementation leads to several shortcomings, such as higher load voltage total harmonic distortion (THD), large tracking error of control objectives, and variable switching frequency. Furthermore, the fourth leg operates with a higher average switching frequency than the other three main legs, which complicates the $LC$ filter design. To overcome the aforementioned shortcomings, a new predictive duty cycle control is proposed in this article. Extensive experimental results are shown to verify that the proposed method can achieve enhanced steady-state performance with significant reduced load voltage THD and tracking error under different operating conditions, such as nonlinear single-/three-phase unbalanced load and single-phase short-circuit fault, while preserving the fast dynamic performance of FCS-MPC. Steady-state and dynamic performances of the proposed method are compared to FCS-MPC and double-vector MPC approaches to validate the superiority of the proposed scheme.