A low-complexity evolutionary algorithm for wind farm layout optimization

The wind farm layout determines the power generation ability, and its optimization remains challenging in terms of algorithm complexity and performance. Several greedy-derived and evolutionaryderived algorithms have been proposed to solve or alleviate the layout planning problems. However, greedy algorithms are efficient in finding the optimal layout but their exploration ability is poor. In contrast, evolutionary algorithms own better global search ability but require huge computational effort to find the optimal layout. Considering that all wind turbines are deployed in the same flat area (i.e. the coordinates of each wind turbine are of dimension 2, and each dimension of all turbines has the same search range), a low-complexity grey wolf optimization technique using a 2-D encoding mechanism (GWOEM) is proposed in this work. which is low-complexity and accelerates search efficiency. The proposed GWOEM effectively combines the advantages of the greedy algorithms and the evolutionary algorithms, and improves the search efficiency while maintaining accuracy. To validate the performance of the proposed GWOEM, a comparative analysis is performed based on two wind scenarios. Simulation results show that the proposed GWOEM is competitive with other state-of-theart techniques. Compared with TDA, ADE, ADE-GRNN, and DEEM, the average power output in the two wind scenarios considered herein increases by 44.41%, 34.45%, 34.97%, and 37.06%, respectively. In terms of execution time, GWOEM is shorter than TDA, ADE, and DEEM except for ADEGRNN which increases by 12.18%. (c) 2023 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).