State Estimation of Three-Phase Four-Conductor Distribution Systems With Real-Time Data From Selective Smart Meters

Yikui Liu, Jie Li, Lei Wu

Research output: Contribution to journalArticlepeer-review

41 Scopus citations

Abstract

Distribution system state estimation (DSSE) has recently been tested and experimentally deployed in some practical distribution networks. Distinct features of distribution systems, such as diverse and unsymmetrical configurations as well as limited real-time measurements, prohibit the direct application of mature state estimation methods for transmission systems. Targeting at three-phase four-conductor configured unsymmetrical medium-voltage distribution systems (MDS) with neutral conductors and ground resistances, this paper proposes a weighted least square-based DSSE approach, in which voltages are chosen as state variables and load pseudo measurements of low-voltage distribution systems (LDS) are considered to compensate insufficient real-time measurements in MDS. Both rectangular and polar coordinates are studied, and voltage variables of neutrals and zero-injection phases are eliminated to reduce the scale of the DSSE problem. Moreover, in order to enhance the load pseudo measurement accuracy of LDSs, a clustering and partial least square regression-based load estimation model is proposed to leverage the real-time communication ability of smart meters. Case studies on a modified IEEE 123-bus distribution system with actual smart meter data illustrate the effectiveness of the proposed approaches.

Original languageEnglish
Article number8611128
Pages (from-to)2632-2643
Number of pages12
JournalIEEE Transactions on Power Systems
Volume34
Issue number4
DOIs
StatePublished - Jul 2019

Keywords

  • Distribution system state estimation
  • load estimation
  • partial least square regression
  • three-phase four-conductor

Fingerprint

Dive into the research topics of 'State Estimation of Three-Phase Four-Conductor Distribution Systems With Real-Time Data From Selective Smart Meters'. Together they form a unique fingerprint.

Cite this