Scholarlink

<p>To address the limitations of the artificial rabbit optimization (ARO) algorithm in handling complex and large-scale optimization problems, this paper proposes an enhanced multi-strategy algorithm—MARO. The proposed algorithm integrates several strategies, including elite opposition-based learning, Levy flight, random walk, and adaptive weight factors, to significantly improve the search capability and convergence accuracy of ARO. First, the algorithm applies the elite opposition-based learning strategy in each iteration to generate opposite solutions, which enriches population diversity and enhances global search capacity, effectively preventing the algorithm from becoming trapped in local optima. Second, it incorporates Levy flight and random walk strategies to increase search randomness during the exploration phase, thus improving the probability of finding the global optimum. Finally, the adaptive weight factor is introduced during the exploitation phase to dynamically adjust the search direction based on different stages, achieving a balance between global exploration and local exploitation. Experimental results demonstrate that MARO consistently outperforms ARO and other popular algorithms across multiple international benchmark test suites. On the CEC2005 test functions, MARO shows significant improvements over CMAES and LSHADE on 9 functions. For the CEC2017 suite, MARO achieves the best average ranking on nearly three-quarters of the test functions, with an average rank of 1.38, significantly outperforming CMAES (3.48) and LSHADE (3.52). On the CEC2019 suite, MARO ranks first in 70% of the test functions. Statistical significance tests, including the Friedman and Wilcoxon tests, confirm that MARO's p-values are less than 0.05 when compared with benchmark algorithms, validating the algorithm's effectiveness. Moreover, MARO also exhibits outstanding optimization performance in real-world engineering problems such as pressure vessel design and reservoir scheduling optimization. The main contribution of this paper lies in the introduction of an innovative multi-strategy optimization framework, which enhances both exploration and exploitation capabilities. This framework effectively addresses the shortcomings of existing ARO algorithms in handling complex problems, providing a more competitive solution for practical optimization tasks.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-strategy enhanced artificial rabbit optimization algorithm for solving engineering optimization problems

  • 期刊论文

  • 发表日期：

    2025年1月9日

查看全文
  • Ni-ni He,
  • Wen-chuan Wang,
  • Jun Wang

摘要

To address the limitations of the artificial rabbit optimization (ARO) algorithm in handling complex and large-scale optimization problems, this paper proposes an enhanced multi-strategy algorithm—MARO. The proposed algorithm integrates several strategies, including elite opposition-based learning, Levy flight, random walk, and adaptive weight factors, to significantly improve the search capability and convergence accuracy of ARO. First, the algorithm applies the elite opposition-based learning strategy in each iteration to generate opposite solutions, which enriches population diversity and enhances global search capacity, effectively preventing the algorithm from becoming trapped in local optima. Second, it incorporates Levy flight and random walk strategies to increase search randomness during the exploration phase, thus improving the probability of finding the global optimum. Finally, the adaptive weight factor is introduced during the exploitation phase to dynamically adjust the search direction based on different stages, achieving a balance between global exploration and local exploitation. Experimental results demonstrate that MARO consistently outperforms ARO and other popular algorithms across multiple international benchmark test suites. On the CEC2005 test functions, MARO shows significant improvements over CMAES and LSHADE on 9 functions. For the CEC2017 suite, MARO achieves the best average ranking on nearly three-quarters of the test functions, with an average rank of 1.38, significantly outperforming CMAES (3.48) and LSHADE (3.52). On the CEC2019 suite, MARO ranks first in 70% of the test functions. Statistical significance tests, including the Friedman and Wilcoxon tests, confirm that MARO's p-values are less than 0.05 when compared with benchmark algorithms, validating the algorithm's effectiveness. Moreover, MARO also exhibits outstanding optimization performance in real-world engineering problems such as pressure vessel design and reservoir scheduling optimization. The main contribution of this paper lies in the introduction of an innovative multi-strategy optimization framework, which enhances both exploration and exploitation capabilities. This framework effectively addresses the shortcomings of existing ARO algorithms in handling complex problems, providing a more competitive solution for practical optimization tasks.