Experimental Assessment of Finite Control Set - Model Predictive Control Applied to a Dual-Converter-Based Rectifier With a Floating DC link

Abstract

This paper proposes a reduced-computation-burden Finite Control Set - Model Predictive Control (FCS-MPC) applied to a dual-converter-based rectifier with a floating DC link. The main goal of this paper is providing a proof of concept of the discussed system employing the proposed FCS-MPC, highlighting its feasibility, simple multivariable control and straightforward implementation. The proposed FCS-MPC reduces the number of tested vectors from the available 64 to only nine, efficiently controlling grid currents and floating DC-link voltage. To evaluate the performance, steady- and transient-state simulations were carried out to compare the proposed FCS-MPC with the conventional PI-based method. The results indicate that FCS-MPC provides a better dynamic response than the PI-based method. However, its total harmonic distortion (THD) at the same sampling frequency is higher, as the PI-based method benefits from a modulation stage that reduces the current ripple. Additionally, the proposed FCS-MPC shows significantly lower switching losses than the PI-based approach. On the other hand, for the same switching frequency, the proposed FCS-MPC presents a somewhat higher, but similar THD and losses values to the PI-based method. Experimental results further validate the feasibility of the proposed FCS-MPC, reinforcing its potential as an efficient alternative to traditional control strategies in dual-converter-based-rectifier.