Energies (Apr 2024)

Nearest Vector Control Method Applied to an MMC for PV Generation

  • Yad N. Bakir,
  • Santiago de Pablo,
  • Fernando Martinez-Rodrigo,
  • Zaid A. Aljawary,
  • Luis Carlos Herrero-de Lucas

DOI
https://doi.org/10.3390/en17081795
Journal volume & issue
Vol. 17, no. 8
p. 1795

Abstract

Read online

This paper proposes a new and simplified Nearest Vector Control (NVC) modulation technique for a grid-connected photovoltaic (PV) system using a Modular Multilevel Converter (MMC). Compared to the Nearest Level Control (NLC) technique, which defines three independent states for the three phases of medium to large four-wire multilevel converters, NVC offers a more coordinated behavior for three-wire converters. The proposed scheme is easy to implement, and it simplifies the understanding of using vectors when detecting the vector of the converter nearest to a given reference. Because it uses natural coordinates, namely, ab, bc and ca, the proposed method is easier to understand and more useful for further developments. Compared with earlier NVC methods, this approach offers full independence of the number of levels at the converter and it can readily accommodate changes in the number of levels, with no need for lookup tables or artificial coordinate transformations. The proposed NVC method was implemented on a 16-cell MMC used for PV generation and then it was compared to NLC, leading to a smaller and more consistent low-order harmonic distortion, requiring about the same complexity of implementation. Furthermore, in comparison to NLC, when applying the proposed NVC modulation, a behavior more insensitive to changes in the grid voltage was found, the most hazardous odd harmonics from the 5th to the 19th were reduced, and a consistent reduction of about 25 dB was achieved on the 5th and 7th harmonics. The newly proposed method is supported by simulations and experimental results with constant and sharply changing solar irradiance, leaving or removing the 100 Hz component of the MMC circulating currents.

Keywords