Scholarlink

<p>Variational mode decomposition (VMD) for allocating power in hybrid energy storage system (HESS) is often hindered by the adaptability issues in parameter selection, limiting dispatch precision. To address this, a multi-strategy improved crested porcupine optimizer (MSICPO) is proposed to adaptively optimize the VMD parameters. The optimizer integrates Kent chaotic mapping, crisscross algorithm, Cauchy mutation, and adaptive population sizing to enhance global search performance and convergence reliability. A composite fitness function, combining envelope entropy and the Pearson correlation coefficient, guides the optimization process to ensure both distinct modal separation and high signal reconstruction fidelity. Furthermore, a two-stage intelligent allocation framework based on MSICPO-VMD dual decomposition is established to refine power dispatch granularity. Initially, MSICPO-VMD and modal energy analysis partition the HESS power into a primary energy layer and a high-frequency fluctuation layer. Subsequently, the primary energy layer undergoes secondary decomposition, and its components are adaptively assigned to hydrogen or battery storage units according to their complexity, quantified by sample entropy. Simulation results demonstrate that MSICPO achieves significantly better convergence speed and optimization accuracy compared to the original CPO and several commonly used swarm intelligence algorithms. The proposed strategy effectively mitigates the mode aliasing problems of conventional decomposition methods while fully leveraging the complementary characteristics of each storage unit.</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

MSICPO-VMD Driven Two-Stage Energy Management for Wind-Hydrogen Hybrid Systems

  • 期刊论文

  • 发表日期：

    2026年1月15日

查看全文
  • Lei Zhang,
  • Guanyu Liu,
  • Yunpei Zhai,
  • Zhiyuan Feng,
  • Hang Liu,
  • Gong Chen

摘要

Variational mode decomposition (VMD) for allocating power in hybrid energy storage system (HESS) is often hindered by the adaptability issues in parameter selection, limiting dispatch precision. To address this, a multi-strategy improved crested porcupine optimizer (MSICPO) is proposed to adaptively optimize the VMD parameters. The optimizer integrates Kent chaotic mapping, crisscross algorithm, Cauchy mutation, and adaptive population sizing to enhance global search performance and convergence reliability. A composite fitness function, combining envelope entropy and the Pearson correlation coefficient, guides the optimization process to ensure both distinct modal separation and high signal reconstruction fidelity. Furthermore, a two-stage intelligent allocation framework based on MSICPO-VMD dual decomposition is established to refine power dispatch granularity. Initially, MSICPO-VMD and modal energy analysis partition the HESS power into a primary energy layer and a high-frequency fluctuation layer. Subsequently, the primary energy layer undergoes secondary decomposition, and its components are adaptively assigned to hydrogen or battery storage units according to their complexity, quantified by sample entropy. Simulation results demonstrate that MSICPO achieves significantly better convergence speed and optimization accuracy compared to the original CPO and several commonly used swarm intelligence algorithms. The proposed strategy effectively mitigates the mode aliasing problems of conventional decomposition methods while fully leveraging the complementary characteristics of each storage unit.