Zeitschriftenartikel
Parallel Function Optimisation Using Evolutionary Algorithms and Deterministic Neighbourhood Search
Lade...
Volltext URI
Dokumententyp
Text/Journal Article
Datum
2011
Autor:innen
Zeitschriftentitel
ISSN der Zeitschrift
Bandtitel
Quelle
PARS: Parallel-Algorithmen, -Rechnerstrukturen und -Systemsoftware: Vol. 28, No. 1
Verlag
Gesellschaft für Informatik e.V.
Zusammenfassung
Modern computer hardware provides massive computational power by parallelism. However, many of the existing algorithms and frameworks are optimised for sequential execution and are not capable to be parallelised or do not scale well on complex parallel architectures. In our paper, we present a metaheuristic consisting of a parallel Evolutionary Algorithm and a parallel Neighbourhood Search. For the implementation massively parallel GPUs are used. This framework is evaluated on the application of function optimisation.