Randomising block sizes for BlockCopy-based wind farm layout optimisation

Simple item page
dc.contributor.authorMayo, Michaelen_NZ
dc.contributor.authorDaoud, Maisaen_NZ
dc.contributor.authorZheng, Chenen_NZ
dc.contributor.editorLeu, Georgeen_NZ
dc.contributor.editorSingh, Hemant Kumaren_NZ
dc.contributor.editorElsayed, Saberen_NZ
dc.coverage.spatialCanberra, Australiaen_NZ
dc.date.accessioned2016-12-01T01:44:16Z
dc.date.available2016en_NZ
dc.date.available2016-12-01T01:44:16Z
dc.date.issued2016en_NZ
dc.description.abstractThe BlockCopy stochastic local search algorithm is a state-of-the-art optimiser for the Wind Farm Layout Optimisation problem. Unlike many other metaheuristics-based optimisers, BlockCopy requires the specification of only one key parameter, namely a block size. In this paper, we investigate the effect on different block sizes on the optimisation results. Using standard benchmarks for the Wind Farm Layout Optimisation problem, we show that smaller fixed block sizes (relative to overall layout size) produce better optimised layouts than larger fixed block sizes. More interestingly, we also show that randomising the block size parameter results in optimisation performance at the same or a better level than that produced by the best algorithm with a fixed block size. Effectively, this means that the user can ignore the need to tune the block size parameter and simply randomise it instead. Such a strategy results in what is effectively a parameterless, but none-the-less effective, optimisation algorithm for the Wind Farm Layout Optimisation problem.
dc.format.mimetypeapplication/pdf
dc.identifier.citationMayo, M., Daoud, M., & Zheng, C. (2016). Randomising block sizes for BlockCopy-based wind farm layout optimisation. In G. Leu, H. K. Singh, & S. Elsayed (Eds.), Proceedings of the 20th Asia Pacific Symposium on Intelligent and Evolutionary Systems, Canberra, Australia, November 2016 (pp. 277–289). Cham, Switzerland: Springer International Publishing AG. http://doi.org/10.1007/978-3-319-49049-6_20en
dc.identifier.doi10.1007/978-3-319-49049-6_20en_NZ
dc.identifier.isbn978-3-319-49049-6en_NZ
dc.identifier.urihttps://hdl.handle.net/10289/10779
dc.language.isoen
dc.publisherSpringer International Publishing AGen_NZ
dc.relation.isPartOfProceedings of the 20th Asia Pacific Symposium on Intelligent and Evolutionary Systemsen_NZ
dc.rightsThis is an author’s accepted version of an article published in Proceedings of the 20th Asia Pacific Symposium on Intelligent and Evolutionary Systems. © 2016 Springer International Publishing. The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-49049-6_20
dc.sourceIES 2016en_NZ
dc.subjectcomputer scienceen_NZ
dc.subjectwind farm layout optimisationen_NZ
dc.subjectblockcopyen_NZ
dc.subjectlocal searchen_NZ
dc.subjectparameter tuningen_NZ
dc.subjectMachine learning
dc.titleRandomising block sizes for BlockCopy-based wind farm layout optimisationen_NZ
dc.typeConference Contribution
pubs.begin-page277
pubs.elements-id143581
pubs.end-page289
pubs.finish-date2016-11-18en_NZ
pubs.organisational-group/Waikato
pubs.organisational-group/Waikato/FCMS
pubs.organisational-group/Waikato/FCMS/Computer Science
pubs.organisational-group/Waikato/FCMS/Computer Science/ML Group
pubs.place-of-publicationCham, Switzerland
pubs.start-date2016-11-16en_NZ
pubs.volumeProceedings in Adaptation, Learning and Optimization 8en_NZ
uow.verification.statusunverified
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
IES paper.pdf
Size:
382.5 KB
Format:
Adobe Portable Document Format
Description:
Accepted version
Thumbnail Image
Name: IES paper.pdf
Size: 382.5 KB
Kind: Adobe PDF
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Research Commons Deposit Agreement 2016.txt
Size:
263 B
Format:
Unknown data format
Description:
Name: Research Commons Deposit Agreement 2016.txt
Size: 263 B
Kind: Unknown
Collections
Computing and Mathematical Sciences Papers