Scholarlink

<p>The Gradient-Based Optimizer (GBO) algorithm combines population-based and gradient-based techniques to enhance both global and local search operations. Despite its promising framework, GBO faces some limitations such as a tendency to get trapped in suboptimal local solutions and inefficiencies when dealing with large-scale problems, which result in increased computational costs. To overcome these shortcomings, a novel chaotic fast random opposition learning-based GBO algorithm is proposed in this study. More precisely, this algorithm integrates chaotic maps to dynamically adjust key parameters of GBO that enhance its ability based on balance exploration and exploitation. In addition, it employs fast random opposition-based learning strategy to accelerate convergence and avoid local optima by generating opposite solutions. The proposed algorithm enhances the performance of the classical GBO by intelligently initializing the population using a chaotic fast random opposition-based learning approach. Finally, the effectiveness and superiority of the proposed optimization algorithm are validated through experimental evaluations on CEC 2005 and CEC 2022 benchmark functions, four real-world engineering problems, along with a nonlinear model predictive control design for single mobile robot model.</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

Improved gradient-based optimizer algorithm using chaotic fast random opposition-based learning

  • 期刊论文

  • 发表日期：

    2026年6月26日

查看全文
  • Gopi Subramani,
  • Rajkumar Selvaraj,
  • Saravanakumar Thangavel,
  • Selvaraj Palanisamy

摘要

The Gradient-Based Optimizer (GBO) algorithm combines population-based and gradient-based techniques to enhance both global and local search operations. Despite its promising framework, GBO faces some limitations such as a tendency to get trapped in suboptimal local solutions and inefficiencies when dealing with large-scale problems, which result in increased computational costs. To overcome these shortcomings, a novel chaotic fast random opposition learning-based GBO algorithm is proposed in this study. More precisely, this algorithm integrates chaotic maps to dynamically adjust key parameters of GBO that enhance its ability based on balance exploration and exploitation. In addition, it employs fast random opposition-based learning strategy to accelerate convergence and avoid local optima by generating opposite solutions. The proposed algorithm enhances the performance of the classical GBO by intelligently initializing the population using a chaotic fast random opposition-based learning approach. Finally, the effectiveness and superiority of the proposed optimization algorithm are validated through experimental evaluations on CEC 2005 and CEC 2022 benchmark functions, four real-world engineering problems, along with a nonlinear model predictive control design for single mobile robot model.