An enhanced crossover strategy for the artificial lemming algorithm for engineering design optimization

To address the limitations of the artificial lemming algorithm (ALA) in convergence accuracy and premature convergence, this paper proposes an enhanced variant, the cross-strategy-integrated artificial lemming algorithm (CALA). Specifically, the CALA integrates a linear inertia weight strategy to balance exploration and exploitation, a historical best-guided strategy to enhance local search, and a crossover strategy to maintain population diversity. The proposed algorithm was evaluated on the IEEE CEC2017 and CEC2022 benchmark suites, achieving minimum Friedman mean ranks of 1.37 and 1.5, respectively, and outperforming several state-of-the-art algorithms in terms of accuracy and robustness. Furthermore, the CALA was successfully applied to constrained engineering design problems and a photovoltaic model parameter estimation task, demonstrating its effectiveness and practical applicability.