Submitted:
11 November 2026
Posted:
12 November 2026
You are already at the latest version
Abstract
In some applications of voltage-source inverters (VSI) in UPS systems, a lower switching fre-quency (about 10 kHz) is advisable, e.g., to reduce power losses. The cheapest control solution is Single-Input-Single-Output (SISO) control when only output-voltage measurement is required. However, such a SISO control does not work perfectly at low switching frequencies because of the dynamic characteristics of the measuring trace, which can be modelled in the frequency range up to the output filter's resonant frequency as a delay equal to some switching periods. The lower the frequency, the higher the delay. In the case of the SISO, the traditional solution could be, e.g., the Smith Predictor. But in this case, the accuracy of the control depends strongly on the accuracy of the inverter model. The previously shown results of using the Smith Predictor were not spectacu-lar. The new approach is the application of the Equivalent Input Disturbance (EID) when the control SISO system works independently, and the EID (using the inverter model) just compen-sates the disturbances, decreasing output voltage distortions – in our case, the output current is treated as the independent disturbance. The disadvantage of this method is the requirement of a second measurement trace of the inductor current what increases the VSI cost. Three versions of EID control law were presented – the continuous control based on the VSI continuous model, the discretized continuous control, and the discrete control based on the VSI discrete model. The sim-ulations of the SISO Coefficient Diagram Model (CDM) with the chosen EID version were shown for the standard nonlinear rectifier RC load (Power Factor = 0.7). Finally, the measurements of the experimental VSI model with the chosen CDM&EID control were presented.
Keywords:
voltage source inverter
; single-input single-output control
; coefficient-diagram-method
; equivalent input disturbance
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.