Scholarlink

<p>The accurate estimation of magnetic target states is crucial for geomagnetic detection, national security, and resource exploration. However, conventional methods often suffer from low convergence precision and a tendency to converge to local optima under complex noise conditions. To address these limitations, we developed a magnetic dipole array model based on a single three-component magnetic sensor, which decomposes the state estimation problem into two subproblems: parameter optimization and magnetic moment solution. For parameter optimization, we propose an Improved Arctic Puffin Optimization (IAPO) algorithm, which integrates a multi-population fusion evolution strategy to enhance global search capability and an adaptive Lévy flight strategy to balance exploration and exploitation. To further improve efficiency, the state at the moment of maximum magnetic field modulus is estimated, thereby narrowing the parameter search space. For the magnetic moment solution, we introduce a Conjugate Gradient Least Squares method regularized by Generalized Cross-Validation (CGLS-GCV), which significantly improves inversion stability. By combining IAPO and CGLS-GCV, we construct a novel state estimation method, termed IAPO-CGLS-GCV. Extensive experiments demonstrate that IAPO significantly outperforms seven state-of-the-art metaheuristic algorithms proposed since 2024 on twelve CEC2022 benchmark functions, as validated through qualitative and quantitative assessments, statistical tests, and ablation studies. Furthermore, numerical simulations and ship-scale model experiments confirm the effectiveness and advancement of the proposed method in estimating magnetic anomaly targets, offering a new pathway for high-noise magnetic target detection.</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

Large-scale magnetic anomaly target state estimation method based on IAPO-CGLS-GCV algorithm

  • 期刊论文

  • 发表日期：

    2025年11月19日

查看全文
  • Binjie Lu,
  • Xiaobing Zhang

摘要

The accurate estimation of magnetic target states is crucial for geomagnetic detection, national security, and resource exploration. However, conventional methods often suffer from low convergence precision and a tendency to converge to local optima under complex noise conditions. To address these limitations, we developed a magnetic dipole array model based on a single three-component magnetic sensor, which decomposes the state estimation problem into two subproblems: parameter optimization and magnetic moment solution. For parameter optimization, we propose an Improved Arctic Puffin Optimization (IAPO) algorithm, which integrates a multi-population fusion evolution strategy to enhance global search capability and an adaptive Lévy flight strategy to balance exploration and exploitation. To further improve efficiency, the state at the moment of maximum magnetic field modulus is estimated, thereby narrowing the parameter search space. For the magnetic moment solution, we introduce a Conjugate Gradient Least Squares method regularized by Generalized Cross-Validation (CGLS-GCV), which significantly improves inversion stability. By combining IAPO and CGLS-GCV, we construct a novel state estimation method, termed IAPO-CGLS-GCV. Extensive experiments demonstrate that IAPO significantly outperforms seven state-of-the-art metaheuristic algorithms proposed since 2024 on twelve CEC2022 benchmark functions, as validated through qualitative and quantitative assessments, statistical tests, and ablation studies. Furthermore, numerical simulations and ship-scale model experiments confirm the effectiveness and advancement of the proposed method in estimating magnetic anomaly targets, offering a new pathway for high-noise magnetic target detection.