A chaotic local search-based LSHADE with enhanced memory storage mechanism for wind farm layout optimization

Abstract

The search for clean energy alternatives to fossil fuels has been a major effort by researchers all over the world. Wind energy is one of the most optimal choices because of its cleanliness and renewability. However, the existence of the wake effect leads to a decrease in conversion efficiency. Finding the best wind turbine layout has become an important factor in the wind power generation system. Inspired by the excellent optimization capability of meta-heuristic algorithms, they are increasingly applied to solve complex constraints and design objectives in the wind farm layout optimization problems. It is reported that LSHADE, which is an advanced variant of differential evolution, provides a more efficient configuration of wind turbines than other meta-heuristic algorithms. This motivates us to conduct research in this direction and design an effective meta-heuristic algorithm with a chaotic local search strategy and an enhanced memory storage mechanism, which contributes to the reduction of global carbon emissions. The proposed new algorithm is called CLSHADE. The validity of the proposed algorithm is verified by the simulation of different constraints and wind field distribution profiles. Compared to four state-of-the-art meta-heuristic algorithms, the average conversion rate of the proposed algorithm is 92.87%, 89.13%, and 96.86% for three wind distribution profiles, respectively. The results show that the proposed algorithm has superiorities and effectiveness in wind farm layout optimization.