Scholarlink

<p>Innovations in data engineering, computational resource availability, and novel mathematical modeling approaches encourage research in optimization to provide promising solutions for a range of challenges in diverse fields. The need and advancements have thus led to researchers proposing evolutionary algorithms (EAs) to solve optimization challenges. However, specifically for EAs, mechanisms to address the global and local search strategies are vital for implementation and solution. To this end, we propose a formalized EA focusing on adaptive switching between global search and local search while leveraging the <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\upvarepsilon \)</EquationSource> <EquationSource Format="MATHML"><math> <mi mathvariant="normal">ε</mi> </math></EquationSource> </InlineEquation>-constraint method for constraint handling. The proposed algorithm uses a novel switching parameter to heuristically switch between global search and local search based on objective function value. Diverse optimization test suites, including standard unconstrained test functions and CEC 2011 real world optimization problems, are used to compare the results of the proposed algorithm to other well-known EAs. In addition, CEC 2020 real world single objective constrained optimization problems are included in the experimental evaluation to prove the effectiveness of the proposed algorithm.</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

Adaptive switching heuristic based evolutionary algorithm: design, validation and comparison

  • 期刊论文

  • 发表日期：

    2026年1月27日

查看全文
  • Tejas M Vala,
  • Vipul N Rajput,
  • Kevin Joshi,
  • Kartik S Pandya,
  • Santosh C Vora

摘要

Innovations in data engineering, computational resource availability, and novel mathematical modeling approaches encourage research in optimization to provide promising solutions for a range of challenges in diverse fields. The need and advancements have thus led to researchers proposing evolutionary algorithms (EAs) to solve optimization challenges. However, specifically for EAs, mechanisms to address the global and local search strategies are vital for implementation and solution. To this end, we propose a formalized EA focusing on adaptive switching between global search and local search while leveraging the \(\upvarepsilon \) ε -constraint method for constraint handling. The proposed algorithm uses a novel switching parameter to heuristically switch between global search and local search based on objective function value. Diverse optimization test suites, including standard unconstrained test functions and CEC 2011 real world optimization problems, are used to compare the results of the proposed algorithm to other well-known EAs. In addition, CEC 2020 real world single objective constrained optimization problems are included in the experimental evaluation to prove the effectiveness of the proposed algorithm.