Scholarlink

<p>Parameters identification of photovoltaic models and engineering problems with constraints are recognized as complex optimization tasks in real-world applications. In this paper, an upgraded variant of Dung Beetle Optimizer (MSDBO) is proposed. MSDBO incorporates three key strategies. Firstly, dynamic population size variation dynamically adjusts the population size to balance global exploration in the early stages and local exploitation in the later stages. Secondly, forced ranking strategy prioritizes higher-quality solutions, accelerating the convergence speed by focusing on the most promising candidates. Finally, boundary control strategy ensures that candidate solutions remain within the feasible region of the search space, improving the robustness of the algorithm and preventing infeasible solutions. These strategies collectively contribute to the improved performance of MSDBO in terms of convergence speed, solution accuracy, and computational efficiency. By leveraging information from high-performing individuals, the algorithm is guided toward more effective solutions. To verify the performance of MSDBO, it was compared with other advanced optimization algorithms on the CEC 2014, CEC 2019, and CEC 2011 test suites. In addition, MSDBO was applied to two types of real-world optimization problems: several specific engineering problems with constraints and parameters identification of photovoltaic models. The experimental results indicate that MSDBO performs competitively compared to other algorithms. Statistical analysis, including the Friedman rank test, reveals that MSDBO ranks first among all tested algorithms, suggesting its effectiveness in both benchmark and real-world applications.</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 boosted dung beetle optimizer and its applications for photovoltaic models and engineering applications

  • 期刊论文

  • 发表日期：

    2025年9月22日

查看全文
  • Xiaoyu Liu,
  • Qingke Zhang,
  • Junqing Li,
  • Huaxiang Zhang

摘要

Parameters identification of photovoltaic models and engineering problems with constraints are recognized as complex optimization tasks in real-world applications. In this paper, an upgraded variant of Dung Beetle Optimizer (MSDBO) is proposed. MSDBO incorporates three key strategies. Firstly, dynamic population size variation dynamically adjusts the population size to balance global exploration in the early stages and local exploitation in the later stages. Secondly, forced ranking strategy prioritizes higher-quality solutions, accelerating the convergence speed by focusing on the most promising candidates. Finally, boundary control strategy ensures that candidate solutions remain within the feasible region of the search space, improving the robustness of the algorithm and preventing infeasible solutions. These strategies collectively contribute to the improved performance of MSDBO in terms of convergence speed, solution accuracy, and computational efficiency. By leveraging information from high-performing individuals, the algorithm is guided toward more effective solutions. To verify the performance of MSDBO, it was compared with other advanced optimization algorithms on the CEC 2014, CEC 2019, and CEC 2011 test suites. In addition, MSDBO was applied to two types of real-world optimization problems: several specific engineering problems with constraints and parameters identification of photovoltaic models. The experimental results indicate that MSDBO performs competitively compared to other algorithms. Statistical analysis, including the Friedman rank test, reveals that MSDBO ranks first among all tested algorithms, suggesting its effectiveness in both benchmark and real-world applications.