A novel adaptive genetic algorithm for wine farm layout optimization

Feng Liu, Zhifang Wang

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    Abstract

    In this paper we propose an adaptive genetic algorithm (AGA) to optimize wind farm layout in order to achieve a highest aggregate wind power conversion efficiency with the presence of wake effect impacts. Instead of using random 'crossover' in each iteration as in conventional GA, our proposed adaptive algorithm introduces novel relocation of 'bad' turbines so that turbines experiencing worst wake effect impacts in a layout will be relocated to some new and more efficient positions. Therefore each iteration of the AGA can more effectively improve the aggregate wind power conversion efficiency and greatly accelerate the algorithm convergence. We experiment the proposed AGA and compare its performance with conventional GA based on a number of scenarios such as multi-directional wind distribution, offshore and inland wind distribution, with sparse and crowded farm settings. Numerical results verify the effectiveness of the proposed AGA algorithm which is able to locate an optimal layout at a much faster convergence speed and achieve a higher aggregate wind power output from a farm.

    Original languageEnglish
    Title of host publication2017 North American Power Symposium, NAPS 2017
    ISBN (Electronic)9781538626993
    DOIs
    StatePublished - 13 Nov 2017
    Event2017 North American Power Symposium, NAPS 2017 - Morgantown, United States
    Duration: 17 Sep 201719 Sep 2017

    Publication series

    Name2017 North American Power Symposium, NAPS 2017

    Conference

    Conference2017 North American Power Symposium, NAPS 2017
    Country/TerritoryUnited States
    CityMorgantown
    Period17/09/1719/09/17

    Keywords

    • genetic algorithm
    • wake effect
    • wind farm layout optimization problem (WFLOP)

    Fingerprint

    Dive into the research topics of 'A novel adaptive genetic algorithm for wine farm layout optimization'. Together they form a unique fingerprint.

    Cite this