Multi-strategy Enhanced Black Kite Algorithm for Constrained Engineering Optimization Problems

To overcome the premature convergence issue and enhance both convergence speed and accuracy of the Black Kite Algorithm (BKA), this paper proposes a multi-strategy enhanced Black Kite Algorithm (MBKA). Firstly, the Lens Imaging Opposition-Based Learning (LOBL) strategy is used to mutate the leader to prevent premature convergence of the algorithm; Secondly, in the attack phase, the position update conditions are changed and incorporating information disparity in early iterations and leader guidance in later iterations to improve the algorithm's convergence accuracy; Finally, in the migration phase, changing the position updating method and corresponding conditions to accelerate the algorithm's convergence speed and accuracy. Based on 16 benchmark functions and 7 comparison algorithms, the effectiveness of each improved strategy of MBKA is verified, and the convergence accuracy, convergence behavior, and statistical test results are compared and analyzed. Experimental results show that MBKA ranks first in the average ranking, and the p-value of the Wilcoxon test confirms that MBKA has a significant difference from other algorithms. Moreover, when examined on three engineering design problems, MBKA achieved a superior average solution compared to the original BKA. These comprehensive results confirm that MBKA offers excellent convergence characteristics and strong robustness.