Scholarlink

<p>In smart warehousing systems, four-way shuttles (FWS) have emerged as core equipment due to their cross-aisle mobility and high flexibility. Their configuration directly determines system cost and operational efficiency. However, traditional configuration methods struggle to effectively balance these two objectives. To address the insufficient evaluation accuracy caused by parameter oversimplification in traditional models, this study constructs a kinematics-based operation cycle time model. By integrating queueing theory to derive a system congestion attenuation mechanism, the model accurately quantifies the nonlinear impacts of rack layout and multi-shuttle collaborative operations on system efficiency. Furthermore, considering the characteristics of "discrete variables and strong constraint coupling" in configuration optimization, a domain-customized improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) is proposed. This algorithm introduces adaptive crossover and mutation operators, and designs an elite neighborhood search based on physical structures, along with a hierarchical unidirectional constraint repair mechanism of “capacity-efficiency-proportion”. This effectively overcomes the defects of conventional algorithms, such as low feasible solution generation rates under strong constraints and susceptibility to cyclic dependency oscillations. Verifications through multi-scenario engineering cases demonstrate that, under an independently tuned environment, the proposed algorithm significantly outperforms advanced baseline algorithms (e.g., CCMO) in metrics like Hypervolume (HV). Moreover, it exhibits excellent robustness when facing high congestion and cost fluctuations. This research provides theoretical support and a quantitative decision-making basis for the configuration optimization of FWS systems.</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 NSGA-II algorithm and multi-objective trade-off for configuration optimization of four-way shuttle systems

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月3日

查看全文
  • Xiaoguang Liu,
  • Jie Yuwen,
  • Mingliang Yang,
  • Xijun Xu,
  • Qing Dong,
  • Heng Yang,
  • Keyuan Zhao

摘要

In smart warehousing systems, four-way shuttles (FWS) have emerged as core equipment due to their cross-aisle mobility and high flexibility. Their configuration directly determines system cost and operational efficiency. However, traditional configuration methods struggle to effectively balance these two objectives. To address the insufficient evaluation accuracy caused by parameter oversimplification in traditional models, this study constructs a kinematics-based operation cycle time model. By integrating queueing theory to derive a system congestion attenuation mechanism, the model accurately quantifies the nonlinear impacts of rack layout and multi-shuttle collaborative operations on system efficiency. Furthermore, considering the characteristics of "discrete variables and strong constraint coupling" in configuration optimization, a domain-customized improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) is proposed. This algorithm introduces adaptive crossover and mutation operators, and designs an elite neighborhood search based on physical structures, along with a hierarchical unidirectional constraint repair mechanism of “capacity-efficiency-proportion”. This effectively overcomes the defects of conventional algorithms, such as low feasible solution generation rates under strong constraints and susceptibility to cyclic dependency oscillations. Verifications through multi-scenario engineering cases demonstrate that, under an independently tuned environment, the proposed algorithm significantly outperforms advanced baseline algorithms (e.g., CCMO) in metrics like Hypervolume (HV). Moreover, it exhibits excellent robustness when facing high congestion and cost fluctuations. This research provides theoretical support and a quantitative decision-making basis for the configuration optimization of FWS systems.