Quadratic Integration-exploited Model Predictive Current Control (QI-MPCC)-based Flying-Capacitor-Clamped Multilevel Converter (FCCMC)

Abstract

The flying-capacitor-clamped multilevel converter (FCCMC) has inherent converter-leg redundant switching combinations per reference multilevel (intermediate levels) voltage due to its flying capacitors clamped to serially-connected switches. Also, it has the least hardware and control complexity among multilevel converters. Hence, FCCMC is the most prominent among multilevel converters in several-hundred-kVA and low-nominal-DC-voltage power applications. Compared to the classic linearized closed-loop controls, the model predictive current control (MPCC) provides improved closed-loop control performance and achieves multiple linear/nonlinear control objectives parallelly through a discrete mathematical model-based predicted cost-function optimization approach utilizing a finite switching combination set based on the space vector control. Hence, MPCC further enhances the FCCMC's advantages in the above DC-voltage and power ranges. But, the ordinary MPCC predicts the cost functions based on the implicit Euler method. It yields inferior closed-loop control performance unless the sampling rate is impractically high enough because the implicit Euler method is order-one accurate. This paper proposes a new MPCC methodology exploiting the quadratic integration (QI) method, QI-MPCC, for FCCMC. The QI method has an order-four accuracy by utilizing three collocation points in the cost function prediction. Therefore, it significantly improves the closed-loop control performance of MPCC-based FCCMC without requesting an impractically high sampling rate. Simulation results demonstrate the steady-state and dynamic performance of the proposed methodology.