Scholarlink

<p>To address the critical bottleneck of logistics in intelligent manufacturing, this paper investigates the Multi-Objective Flexible Job-shop Scheduling Problem integrated with Automated Guided Vehicles (AGVs). We propose an enhanced knowledge-driven NSGA-II algorithm (EK-NSGA-II) to simultaneously minimize makespan, machine workload imbalance, and AGV load fluctuations. The methodology integrates three innovations: a three-layer chromosome encoding for synchronized resource allocation, a normalized adaptive crossover mechanism for exploration-exploitation balance, and a history-elite-knowledge micro-perturbation local search (EKMPLS) to mitigate “search blindness.” EK-NSGA-II was rigorously evaluated on 30 benchmark instances from Bilge &amp; Ulusoy and Deroussi &amp; Norre datasets, quantitative results demonstrate significant performance leaps: ablation studies show that removing the EKMPLS module increases the Comprehensive Deterioration Rate by 11.6%. Comparative experiments against standard NSGA-II reveal that EK-NSGA-II reduces average machine load by 50.0% and makespan by 12.1%. Statistical validation using the Wilcoxon rank-sum test (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(p &lt; 0.05\)</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mi>p</mi> <mo>&lt;</mo> <mn>0.05</mn> </mrow> </math></EquationSource> </InlineEquation>) confirms the robustness and superior convergence of the proposed approach. These findings validate that distilling latent Pareto patterns from evolutionary memory effectively optimizes dual-resource load balancing in complex integrated scheduling environments.</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

EK-NSGA-II: Elite-Knowledge and Cooperative Selection for Multi-objective FJSP-AGV Optimization

  • 期刊论文

  • 发表日期：

    2026年5月3日

查看全文
  • Ziting Zhang,
  • Ye Tian,
  • Yaxuan Zhang

摘要

To address the critical bottleneck of logistics in intelligent manufacturing, this paper investigates the Multi-Objective Flexible Job-shop Scheduling Problem integrated with Automated Guided Vehicles (AGVs). We propose an enhanced knowledge-driven NSGA-II algorithm (EK-NSGA-II) to simultaneously minimize makespan, machine workload imbalance, and AGV load fluctuations. The methodology integrates three innovations: a three-layer chromosome encoding for synchronized resource allocation, a normalized adaptive crossover mechanism for exploration-exploitation balance, and a history-elite-knowledge micro-perturbation local search (EKMPLS) to mitigate “search blindness.” EK-NSGA-II was rigorously evaluated on 30 benchmark instances from Bilge & Ulusoy and Deroussi & Norre datasets, quantitative results demonstrate significant performance leaps: ablation studies show that removing the EKMPLS module increases the Comprehensive Deterioration Rate by 11.6%. Comparative experiments against standard NSGA-II reveal that EK-NSGA-II reduces average machine load by 50.0% and makespan by 12.1%. Statistical validation using the Wilcoxon rank-sum test ( \(p < 0.05\) p < 0.05 ) confirms the robustness and superior convergence of the proposed approach. These findings validate that distilling latent Pareto patterns from evolutionary memory effectively optimizes dual-resource load balancing in complex integrated scheduling environments.