Improved power flow robustness for personal computers

Daniel Tylavsky, Leslie F. Jarriel, Peter E. Crouch, Rambabu Adapa

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations

Abstract

Mine electrical power-flow solutions are often obtained using personal computers. Because of the limited precision used by the compilers on these machines, power-flow solution procedures may diverge or possibly oscillate until the iteration limit is reached, even though an operable solution exists. Two methods to create virtually extended-precision calculations in full Newton-Raphson and decoupled power-flow algorithms without the attendant slow down associated with full double-precision codes are presented. It is shown that insufficient precision for convergence of the IEEE 118 bus system occurs near the single precision limit of a PC- or VAX-class minicomputer, the number of bits required for convergence can be simply and quickly estimated, and the most critical routine in which precision generally needs to be increased is the mismatch routine. Simulation results are discussed.

Original languageEnglish (US)
Title of host publicationConference Record - IAS Annual Meeting (IEEE Industry Applications Society)
PublisherPubl by IEEE
Pages1401-1407
Number of pages7
Editionpt 2
ISBN (Print)0879425539
StatePublished - Dec 1 1990
Event1990 IEEE Industry Applications Society Annual Meeting - IAS-25 Part 2 (of 2) - Seattle, WA, USA
Duration: Oct 7 1990Oct 12 1990

Publication series

NameConference Record - IAS Annual Meeting (IEEE Industry Applications Society)
Numberpt 2
ISSN (Print)0160-8592

Other

Other1990 IEEE Industry Applications Society Annual Meeting - IAS-25 Part 2 (of 2)
CitySeattle, WA, USA
Period10/7/9010/12/90

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Industrial and Manufacturing Engineering
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Improved power flow robustness for personal computers'. Together they form a unique fingerprint.

Cite this