Scholarlink

<p>The Butterfly Optimization Algorithm is a new meta-heuristic approach for solving optimization problems. However, it faces some challenges when dealing with complex cases. The Multi-strategy Butterfly Optimization Algorithm (MSBOA) is proposed to overcome these shortcomings. In MSBOA, this paper significantly enhances the uniformity of the population by introducing an improved tent chaotic optimization strategy, balances the exploitation and exploration capabilities through an adaptive gbest-guided mechanism, efficiently searches for the unknown region, and avoids premature convergence by utilizing a refraction-based learning strategy. we demonstrate the global convergence of the MSBOA using the Markov chain. In experiments on the 20-dimensional CEC2020, 20-dimensional CEC2022, and 50-dimensional CEC2017 benchmarks, MSBOA outperforms recent advanced optimization algorithms in 28.94% of cases, is comparable in 64.02%, and underperforms in only 7.03%. Finally, the effectiveness of MSBOA in the optimization of complex systems is further verified by applying it to state estimation of sandwich systems and solving high-dimension observer gain matrices.</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 butterfly optimization algorithm for state estimation in sandwich systems

  • 期刊论文

  • 发表日期：

    2025年8月19日

查看全文
  • Xufeng Liu,
  • Zupeng Zhou,
  • Yongquan Zhou

摘要

The Butterfly Optimization Algorithm is a new meta-heuristic approach for solving optimization problems. However, it faces some challenges when dealing with complex cases. The Multi-strategy Butterfly Optimization Algorithm (MSBOA) is proposed to overcome these shortcomings. In MSBOA, this paper significantly enhances the uniformity of the population by introducing an improved tent chaotic optimization strategy, balances the exploitation and exploration capabilities through an adaptive gbest-guided mechanism, efficiently searches for the unknown region, and avoids premature convergence by utilizing a refraction-based learning strategy. we demonstrate the global convergence of the MSBOA using the Markov chain. In experiments on the 20-dimensional CEC2020, 20-dimensional CEC2022, and 50-dimensional CEC2017 benchmarks, MSBOA outperforms recent advanced optimization algorithms in 28.94% of cases, is comparable in 64.02%, and underperforms in only 7.03%. Finally, the effectiveness of MSBOA in the optimization of complex systems is further verified by applying it to state estimation of sandwich systems and solving high-dimension observer gain matrices.