Parallel processing in power systems computation

Abstract

An attempt is made to define the state of the art and identify the most fertile grounds for future research in parallel processing as applied to power system computation. The analysis is confined to examining the application of parallel computers to specific power system problems. The state of the art is summarized in an attempt to capture the developing approach of matching algorithms to architectures. Research problems including neural network use and genetic algorithms all involve the central challenge of matching highly parallelized algorithms to parallel/distributed architectures.<<ETX>>