Scholarlink

<p>In this study, a two-stage hierarchical evolutionary algorithm based on multi-population (MPTSHE) is proposed to solve multi-objective distributed heterogeneous welding flow shop scheduling problems (DHWSP). The algorithm consists of two stages: in the first stage, each population evolves independently to enhance convergence; in the second stage, the populations co-evolve to improve diversity. Additionally, five heuristic rules are designed to generate high-quality initial populations, and two knowledge-based neighborhood structures are constructed using problem characteristics to accelerate the convergence of the populations. Furthermore, a diversity search mode is developed to identify elite populations, which further enhances convergence and diversity of the populations. Finally, a large number of experiments are carried out on 20 DHWSP cases in the latest literatures, and the proposed MPTSHE is compared with the algorithms from other literatures. The experimental results indicate that MPTSHE outperforms existing algorithms in terms of comprehensive performance metrics, demonstrating superior efficiency.</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

Two-stage hierarchical evolutionary algorithm based on multi-population for multi-objective distributed heterogeneous welding flow shop scheduling

  • 期刊论文

  • 发表日期：

    2025年7月31日

查看全文
  • Liangcai Xia,
  • Shijun Chen

摘要

In this study, a two-stage hierarchical evolutionary algorithm based on multi-population (MPTSHE) is proposed to solve multi-objective distributed heterogeneous welding flow shop scheduling problems (DHWSP). The algorithm consists of two stages: in the first stage, each population evolves independently to enhance convergence; in the second stage, the populations co-evolve to improve diversity. Additionally, five heuristic rules are designed to generate high-quality initial populations, and two knowledge-based neighborhood structures are constructed using problem characteristics to accelerate the convergence of the populations. Furthermore, a diversity search mode is developed to identify elite populations, which further enhances convergence and diversity of the populations. Finally, a large number of experiments are carried out on 20 DHWSP cases in the latest literatures, and the proposed MPTSHE is compared with the algorithms from other literatures. The experimental results indicate that MPTSHE outperforms existing algorithms in terms of comprehensive performance metrics, demonstrating superior efficiency.