返回
A surrogate-assisted evolutionary algorithm with an adaptive sample selection strategy for wind farm layout optimization
DOI:10.1080/15435075.2022.2131433.png)
摘要
En 中文
When using evolutionary algorithms to optimize wind farm layouts, it is computationally expensive to use computational fluid dynamics (CFD) model to evaluate different wind farm layouts. To alleviate the difficulty, surrogate models are used to replace CFD model. Such algorithms are called surrogate-assisted evolutionary algorithms (SAEAs). In existing SAEAs, constructing a global surrogate for high-dimensional problems is not reliable due to the curse of dimensionality. Additionally, a local surrogate model constructed with a fixed number of samples has limitations. To address these problems, an SAEA combining global exploration and local exploitation is proposed. The global exploration is based on an adaptive surrogate-assisted particle swarm optimization algorithm, while the local exploitation is based on a local RBF-assisted simulated annealing algorithm. In building a local RBF model, according to the current function evaluation, a certain amount of outstanding samples are selected from the data set. 10 benchmark functions with different modalities and dimensions have been tested to verify the effectiveness of the proposed algorithm. The experimental results show that the proposed algorithm performs well for most benchmark functions. The proposed algorithm was applied to optimize wind farm layout, and achieved an 8.3% increase in annual net power generation.
Keyword:
Wind farm layout
surrogate-assisted evolutionary algorithm
radial basis function
simulated annealing algorithm
particle swarm optimization
期刊
IF:
3.1
论文数:
2.4K
被引数:
3.9K
机构
暂无机构信息
引用论文
New approach on optimization in placement of wind turbines within wind farm by genetic algorithms
RENEWABLE ENERGY
IF9.1
Differential Action between Schisandrin A and Schisandrin B in Eliciting an Anti-Inflammatory Action: The Depletion of Reduced Glutathione and the Induction of an Antioxidant Response
PLOS ONE
IF0
A review on design of experiments and surrogate models in aircraft real-time and many-query aerodynamic analyses飞机实时和多查询气动分析中实验和代理模型的设计综述

