Back to results

University of Arkansas

A Virtual Space Vectors based Model Predictive Control for Three-Level Converters

Abstract

dc:description.abstract

<p>Three-phase three-level (3-L) voltage source converters (VSC), e.g., neutral-point clamped (NPC) converters, T-type converters, etc., have been deemed to be suitable for a wide range of medium- to high-power applications in microgrids (MGs) and bulk power systems. Compared to their two-level (2-L) counterparts, adopting 3-L VSCs in the MG applications not only reduces the voltage stress across the power semiconductor devices, which allows achieving higher voltage levels, but also improves the quality of the converter output waveforms, which further leads to considerably smaller output ac passive filters.</p> <p>Various control strategies have been proposed and implemented for 3-L VSCs. Among all the existing control methods, finite-control-set model predictive control (FCS-MPC) has been extensively investigated and applied due to its simple and intuitive design, fast-dynamic response and robustness against parameter uncertainties. However, to implement an FCS-MPC on a 3-L VSC, a multi-objective cost function, which consists of a term dedicated specifically to control the dc-link capacitor voltages such that the neutral-point voltage (NP-V) oscillations are minimized, must be designed. Nevertheless, selecting proper weighting factors for the multiple control objectives is difficult and time consuming. Additionally, adding the dc-link capacitor voltages balancing term to the cost function distributes the controller effort among different control targets, which severely impacts the primary goal of the FCS-MPC. Furthermore, to control the dc-link capacitor voltages, additional sensing circuitries are usually necessary to measure the dc-link capacitor voltages and currents, which consequently increases the system cost, volume and wiring complexity as well as reduces overall reliability.</p> <p>To address all the aforementioned challenges, in this dissertation research, a novel FCS-MPC method using virtual space vectors (VSVs), which do not affect the dc-link capacitor voltages of the 3-L VSCs, was proposed, implemented and validated. The proposed FCS-MPC strategy has the capability to achieve inherent balanced dc-link capacitor voltages. Additionally, the demonstrated control technique not only simplifies the controller design by allowing the use of a simplified cost function, but also improves the quality of the 3-L VSC output waveforms. Furthermore, the execution time of the proposed control algorithm was significantly reduced compared to that of the existing one. Lastly, the proposed FCS-MPC using the VSVs reduces the hardware cost and complexity as the additional dc-link capacitor voltages and current sensors are not required, which further enhances the overall system reliability.</p>

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy in Engineering (PhD)
Level thesis:degree_level
Dissertation
Year dc:date.available
2020

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Alhosaini, Waleed Saad S.
Advisor dc:contributor.advisor
  • Zhao, Yue
Contributors dc:contributor
  • Ang, Simon S.
  • McCann, Roy A.

Subjects

dc:subject × 8

Identifiers

dc:identifier.*
Repository record dc:identifier
https://scholarworks.uark.edu/etd/3593
OAI identifier oai:identifier
oai:scholarworks.uark.edu:etd-5166

Chain of custody

source
Harvested from
University of Arkansas
Base URL
scholarworks.uark.edu/do/oai/
Last updated
2026-07-24
Source record
OAI-PMH GetRecord
citation

Alhosaini, Waleed Saad S.. A Virtual Space Vectors based Model Predictive Control for Three-Level Converters. Dissertation thesis, 2020. https://scholarworks.uark.edu/etd/3593