Scholarlink

<p>Solving complex optimization problems within a reasonable timeframe had become increasingly difficult and challenging in the contemporary era of industrialization. To address this, metaheuristic algorithms (MAs) have emerged as simple yet efficient methods for solving highly non-linear and multi-modal real-world optimization problems, thus gaining attention from researchers belonging to diverse fields of science as well as engineering. In recent years, MAs have witnessed and undergone substantial advancements driven by the incorporation of nature-inspired mechanisms, hybridization with machine learning, adaptive parameter control, and applications in numerous evolving areas such as big data analytics, cloud computing, engineering design, and healthcare. While researchers have shared many great findings related to advancements in MAs in numerous research articles over the past decade, there are still many areas yet to be explored. With these objectives in mind, this paper presents a comprehensive review of metaheuristic algorithms developed and refined over the period of the past 5&#xa0;years, with an attention on both methodological innovations and real-world applications. A structured taxonomy is projected that categorizes algorithms according to their inspiration sources, search strategies, and improvement techniques, allowing systematic comparison and clearer understanding of their evolution. Benchmark functions and engineering design problems are discussed in context to algorithmic performance, while performance assessment techniques are discussed to highlight convergence behaviour, robustness, and scalability. Constructive criticism is also highlighted in the paper to identify methodological limitations and unresolved challenges. The paper finally concludes by outlining open challenges such as balancing exploration and exploitation, improving generalization, and integrating with emerging technologies and proposing promising avenues for the future development and deployment of metaheuristic algorithms.</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

Metaheuristic Algorithms Since 2020: Development, Taxonomy, Analysis, and Applications

  • 期刊论文

  • 发表日期：

    2025年10月17日

查看全文
  • Arunita Das,
  • Rebika Rai,
  • Buddhadev Sasmal,
  • Krishna Gopal Dhal,
  • Ruba Abu Khurma,
  • Ramesh Saha

摘要

Solving complex optimization problems within a reasonable timeframe had become increasingly difficult and challenging in the contemporary era of industrialization. To address this, metaheuristic algorithms (MAs) have emerged as simple yet efficient methods for solving highly non-linear and multi-modal real-world optimization problems, thus gaining attention from researchers belonging to diverse fields of science as well as engineering. In recent years, MAs have witnessed and undergone substantial advancements driven by the incorporation of nature-inspired mechanisms, hybridization with machine learning, adaptive parameter control, and applications in numerous evolving areas such as big data analytics, cloud computing, engineering design, and healthcare. While researchers have shared many great findings related to advancements in MAs in numerous research articles over the past decade, there are still many areas yet to be explored. With these objectives in mind, this paper presents a comprehensive review of metaheuristic algorithms developed and refined over the period of the past 5 years, with an attention on both methodological innovations and real-world applications. A structured taxonomy is projected that categorizes algorithms according to their inspiration sources, search strategies, and improvement techniques, allowing systematic comparison and clearer understanding of their evolution. Benchmark functions and engineering design problems are discussed in context to algorithmic performance, while performance assessment techniques are discussed to highlight convergence behaviour, robustness, and scalability. Constructive criticism is also highlighted in the paper to identify methodological limitations and unresolved challenges. The paper finally concludes by outlining open challenges such as balancing exploration and exploitation, improving generalization, and integrating with emerging technologies and proposing promising avenues for the future development and deployment of metaheuristic algorithms.