Precedence Relationship Performance of an Indirect Matrix Solver

Daniel Tylavsky, Balakrishnan Gopalakrishnan

Research output: Contribution to journalArticle

Abstract

Indirect methods have been eschewed for use on serial processors by the power industry but may be competitive with direct methods for utilizing large scale parallel processors if the number of precedence relationships experienced by both direct and indirect methods are comparable. A new highly parallel truncated accelerated Neumann series preconditioner is presented. Results from using this preconditioner with a Jacobi-like iterative scheme on a 118 and 494 bus system are presented. These results show that the number of precedence relationships experienced by this technique are of the same order of magnitude as the number encountered when using direct methods.

Original languageEnglish (US)
Pages (from-to)412-418
Number of pages7
JournalIEEE Transactions on Power Systems
Volume5
Issue number2
DOIs
StatePublished - May 1990

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Precedence Relationship Performance of an Indirect Matrix Solver'. Together they form a unique fingerprint.

  • Cite this