Scholarlink

<p>The imbalance between exploration and exploitation represents the central challenge in the Hippopotamus Optimization (HO) algorithm. To enhance the global search capability and convergence accuracy of the HO algorithm, we propose the preferred hybrid strategy improved hippopotamus optimization (HSIHO) algorithm and engineering application with electromagnetic data denoising. This paper reviews various improvement strategies for intelligent optimization algorithms and implements a targeted enhancement for the HO algorithm. The proposed preferred HSIHO integrates lens opposite-based learning (LOBL) with adaptive t-distribution perturbation (ATP) strategy. This approach advantages include enhancing population diversity and global exploration capabilities, improving convergence speed and solution quality, dynamically balancing exploration and exploitation, strengthening the robustness of the optimal solution, possessing a strong ability to escape local optima, and featuring a simple mechanism with broad applicability. Comparing the convergence performance of a single strategy with that of the improved hybrid strategy. Experimental results demonstrate that the HSIHO outperforms other intelligent optimization algorithms in terms of convergence speed, computational efficiency, solution accuracy, flexibility, scalability across multiple benchmark function tests, thereby validating both the efficacy of our preferred improvement strategy and the potential applications of this algorithm. Meanwhile, the preferred HSIHO algorithm are used for parameters optimization of variational mode decomposition (VMD) method and long short term memory (LSTM) method, and compared with empirical mode decomposition (EMD) method, VMD method, particle swarm optimization (PSO)-VMD method and probabilistic neural networks (PNN) method, convolutional neural networks (CNN) method, LSTM method in the electromagnetic data denoising. The proposed method can achieve optimal results in multiple parameter indicators. The denoising data is more in line with the characteristics of the effective signal of the electromagnetic method, and the data quality has been significantly improved. Furthermore, the satisfactory performance in the engineering application verifies the effectiveness of the design and optimization method.</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

Preferred hybrid strategy improved hippopotamus optimization algorithm and engineering application with electromagnetic data denoising

  • 期刊论文

  • 发表日期：

    2026年7月2日

查看全文
  • Xian Zhang,
  • Ruyu Wang,
  • Zengjin OuYang,
  • Minjuan Zhong

摘要

The imbalance between exploration and exploitation represents the central challenge in the Hippopotamus Optimization (HO) algorithm. To enhance the global search capability and convergence accuracy of the HO algorithm, we propose the preferred hybrid strategy improved hippopotamus optimization (HSIHO) algorithm and engineering application with electromagnetic data denoising. This paper reviews various improvement strategies for intelligent optimization algorithms and implements a targeted enhancement for the HO algorithm. The proposed preferred HSIHO integrates lens opposite-based learning (LOBL) with adaptive t-distribution perturbation (ATP) strategy. This approach advantages include enhancing population diversity and global exploration capabilities, improving convergence speed and solution quality, dynamically balancing exploration and exploitation, strengthening the robustness of the optimal solution, possessing a strong ability to escape local optima, and featuring a simple mechanism with broad applicability. Comparing the convergence performance of a single strategy with that of the improved hybrid strategy. Experimental results demonstrate that the HSIHO outperforms other intelligent optimization algorithms in terms of convergence speed, computational efficiency, solution accuracy, flexibility, scalability across multiple benchmark function tests, thereby validating both the efficacy of our preferred improvement strategy and the potential applications of this algorithm. Meanwhile, the preferred HSIHO algorithm are used for parameters optimization of variational mode decomposition (VMD) method and long short term memory (LSTM) method, and compared with empirical mode decomposition (EMD) method, VMD method, particle swarm optimization (PSO)-VMD method and probabilistic neural networks (PNN) method, convolutional neural networks (CNN) method, LSTM method in the electromagnetic data denoising. The proposed method can achieve optimal results in multiple parameter indicators. The denoising data is more in line with the characteristics of the effective signal of the electromagnetic method, and the data quality has been significantly improved. Furthermore, the satisfactory performance in the engineering application verifies the effectiveness of the design and optimization method.