Title |
Investigation of Relationship between DC-link Capacitor Condition and Performance of Three-phase AC/DC Converters Controlled by Model Predictive Control Methods |
Authors |
김동연(Dongyeon Kim) ; 김재창(Jae-Chang Kim) ; 곽상신(Sangshin Kwak) |
DOI |
https://doi.org/10.5370/KIEE.2021.70.4.626 |
Keywords |
Three-phase active AC/DC converter; Model predictive control methods; DC-link capacitor conditions |
Abstract |
In this paper, six model predictive control techniques developed so far to be used to operate a three-phase active rectifier are classified. In addition, their input and output electrical performance are compared and analyzed. The six model predictive control techniques are largely classified into voltage-based (VOC) and direct power control (DPC) methods. In addition, VOC and DPC are classified into a basic control method and a modified control method divided into a vector selection method and an offset voltage injection method, respectively. The electrical characteristics of the input and the output side of the three-phase active rectifier operated by these six types of model prediction control algorithms are compared. In particular, the change of characteristics according to the variation of capacitance and equivalent series resistor (ESR) of the output DC link capacitor is investigated and analyzed. As a result, it is confirmed that the performance change of the three-phase active rectifier is greater due to the increase in the ESR than the decrease in the capacitance of the output DC link capacitor. In addition, it is confirmed that in both the basic and modified algorithms, the DPC-based model predictive control algorithm shows better performance in both the input current total harmonic distortion (THD) and the output voltage fluctuation in terms of increase of ESR and reduction of capacitance of capacitor than the VOC-based model predictive control technique. In the capacitor loss, the DPC-based control techniques show higher efficiency than the VOC-based control algorithm. |