Towards realistic designs of wind farm layouts: Application of a novel placement selector approach - Université de Lorraine Access content directly
Journal Articles Energy Conversion and Management Year : 2014

Towards realistic designs of wind farm layouts: Application of a novel placement selector approach

Abstract

The optimal arrangement of wind turbines plays a significant role in obtaining the expected output power from wind farms. This paper addresses challenges related to typical restriction assumptions of turbine arrangement in wind farms with candidate selection approach. An applicable hybrid (quadratic assignment problem-genetic algorithm) evolutionary method with an initial candidate points selection (ICPS) approach is proposed and applied to four case studies to obtain optimal layout designs with maximum efficiency. The current study considers not only the previously utilized indicators found in the literature, such as wake effects, turbine hub height and rotor diameter, but also accounts for additional criteria such as the load-bearing capacity of soil and restrictions regarding the existence of prohibited places as well as varying wind velocities and directions. This is done to make the approach more applicable for realistic cases, and also to incorporate the preferences of expert designers. The results suggest that superior performance is attained with the proposed algorithm compared to previous similar studies. An efficiency improvement of about 3% is achieved for case one, and the algorithm provides reasonable optimal wind farm design layouts for cases two, three, and four where more restrictions exist.

Keywords

No file

Dates and versions

hal-01418374 , version 1 (16-12-2016)

Identifiers

Cite

Omid Rahbari, Majid Vafaeipour, Farivar Fazelpour, Michel Feidt, Marc A. Rosen. Towards realistic designs of wind farm layouts: Application of a novel placement selector approach. Energy Conversion and Management, 2014, 81, pp.242 - 254. ⟨10.1016/j.enconman.2014.02.010⟩. ⟨hal-01418374⟩
45 View
0 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More