Optimization of Wind Farm Layout: A Refinement Method by Random Search

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

4490 Downloads (Pure)

Abstract

Wind farm layout optimization is to find the optimal positions of wind turbines inside a wind farm, so as to maximize and/or minimize a single objective or multiple objectives, while satisfying certain constraints. Most of the works in the literature divide the wind farm into cells in which turbines can be placed, hence, simplifying the problem from with continuous variables to with discrete variables. In this paper, a refinement method, based on continuous formulation, by using random search is proposed to improve the optimization results based on discrete formulations. Two sets of optimization results of a widely studied test case are refined using the proposed method. One set of the results is from a published work using GA based on discrete formulation, the other set is the improved results using authors’ own GA code. Steady improvements are obtained for both sets of results.
Original languageEnglish
Title of host publicationProceedings of the 2013 International Conference on aerodynamics of Offshore Wind Energy Systems and wakes (ICOWES2013)
EditorsWenZhong Shen
PublisherTechnical University of Denmark
Publication date2013
Pages624-633
Publication statusPublished - 2013
EventInternational Conference on aerodynamics of Offshore Wind Energy Systems and wakes (ICOWES 2013) - Lyngby, Denmark
Duration: 17 Jun 201319 Jun 2013

Conference

ConferenceInternational Conference on aerodynamics of Offshore Wind Energy Systems and wakes (ICOWES 2013)
CountryDenmark
CityLyngby
Period17/06/201319/06/2013

Bibliographical note

This work was supported by the international collaboration project (DSF Sagsnr. 10-094544) under Danish Research and Innovation Council for Strategic Research.

Fingerprint Dive into the research topics of 'Optimization of Wind Farm Layout: A Refinement Method by Random Search'. Together they form a unique fingerprint.

Cite this