China Mechanical Engineering ›› 2026, Vol. 37 ›› Issue (7): 1695-1707.DOI: 10.3969/j.issn.1004-132X.2026.07.019

Previous Articles    

Integrated Scheduling for Flexible Job Shops with Heterogeneous Transport Resources

ZHANG Guohui1(), WANG Wendi1, YU Nana1, WU Changjun2, KOU Xiaofei1   

  1. 1.School of Management Engineering,Zhengzhou University of Aeronautics,Zhengzhou,450046
    2.School of Mechanical and Electrical Engineering,Zhengzhou University of Light Industry,Zhengzhou,450002
  • Received:2025-06-18 Online:2026-07-25 Published:2026-08-18
  • Contact: ZHANG Guohui

考虑异构运输资源的柔性作业车间集成调度问题

张国辉1(), 王文迪1, 余娜娜1, 邬昌军2, 寇晓菲1   

  1. 1.郑州航空工业管理学院管理工程学院, 郑州, 450046
    2.郑州轻工业大学机电工程学院, 郑州, 450002
  • 通讯作者: 张国辉
  • 作者简介:张国辉*(通信作者),男,1980年生,教授。研究方向为智能优化算法、车间调度。发表论文65篇。E-mail:zgh09@zua.edu.cn
  • 基金资助:
    国家自然科学基金(52575603);河南省重点研发专项(231111221200);河南省重大科技专项(241100220200);河南省自然科学基金(252300420985)

Abstract:

This paper addresses the integrated scheduling problem of heterogeneous transportation resources—comprising an overhead crane and automated guided vehicles—in flexible job shops. A reinforcement learning-based multi-objective evolutionary algorithm is proposed to minimize both makespan and total energy consumption. Three hybrid initialization strategies are designed to enhance population quality and diversity. Subsequently, a reinforcement learning mechanism is introduced for adaptive control of genetic operator parameters, and a critical-path-based hybrid neighborhood structure is constructed to simultaneously optimize makespan and energy consumption, thereby guiding efficient exploration of high-quality solutions. Finally, ablation and comparative experiments validate the effectiveness and stability of the proposed algorithm in solving the heterogeneous-transportation-resource flexible job shop scheduling problem.

Key words: flexible job shop, heterogeneous transportation resources, reinforcement learning-based NSGA-II, neighborhood structure

摘要:

针对天车与自动引导小车构成的异构运输资源在柔性作业车间中的集成调度问题,以最小化最大完工时间和总能耗为优化目标,提出一种融合强化学习机制的多目标进化算法。设计了三种混合初始化策略以提升种群质量与多样性;引入强化学习以实现遗传算子参数的自适应控制,构建基于关键路径的混合邻域结构以同时优化最大完工时间和总能耗,引导算法高效探索高质量解空间。最后,通过消融实验与对比实验验证了所提算法在求解异构运输资源柔性作业车间调度问题中的有效性与稳定性。

关键词: 柔性作业车间, 异构运输资源, 强化学习增强NSGA-Ⅱ算法, 邻域结构

CLC Number: