Dynamic Dual-Mutation Strategy-Based Grey Wolf Optimizer for Solving Discrete Wind Farm Layout Optimization Problem
Discrete wind farm layout optimization (WFLO) involves discontinuous and highly combinatorial solution spaces, where existing swarm intelligence methods exhibit weak discrete local search and aerodynamic adaptability, limiting power generation and cost efficiency improvements. This study proposes a dynamic dual-mutation strategy-based grey wolf optimizer (DDMS-GWO) to solve discrete WFLO problems. The method utilizes three leading grey wolves for local search, integrating a Hadamard product mutation strategy for elite gene recombination to preserve favorable aerodynamic topologies, and an elite pool-guided differential mutation strategy to enrich population diversity and avoid premature convergence. The two mutation strategies are adaptively scheduled through a dynamic probability, while a first-order exponential smoothing mechanism is employed to stabilize iterative control parameters. DDMS-GWO was validated against nine baseline algorithms across three wind scenarios. Results demonstrate that DDMS-GWO achieves superior normalized cost of energy (CoE), total power output, and overall efficiency. Specifically, in Scenarios 2 and 3, it improves total power output by up to 0.98% and 1.85% while reducing average CoE by up to 1.72% and 2.42%, respectively. Further analyses of convergence, layout topology, and probability-weighted wake exposure confirm that DDMS-GWO generates more coordinated layouts with reduced wake exposure and enhanced aerodynamic rationality.