中国机械工程 ›› 2026, Vol. 37 ›› Issue (10): 2418-2427.DOI: 10.3969/j.issn.1004-132X.2026.10.004

• 低碳设计与全生命周期优化 • 上一篇    下一篇

基于精英型遗传策略和空隙动态填充机制的矩形件优化排样方法

吴电建1,2(), 黄帅康1(), 丁子璨1, 张三强1,2, 曹卫东3   

  1. 1.湖北工业大学农机工程研究设计院, 武汉, 430068
    2.湖北省农机装备智能化工程技术研究中心, 武汉, 430068
    3.河海大学人工智能与自动化学院, 常州, 213200
  • 收稿日期:2025-11-28 出版日期:2026-10-25 发布日期:2026-11-10
  • 通讯作者: 黄帅康
  • 作者简介:吴电建,男,1990年生,博士、讲师。研究方向为农机智能装备设计与优化。E-mail:hbut_wdj@163.com
    黄帅康*(通信作者),男,1998年生,硕士研究生。研究方向为农机智能装备设计与优化。E-mail: 1269884088@qq.com。
  • 基金资助:
    国家重点研发计划(2018YFD0301303);国家自然科学基金(52305532);湖北省技术创新计划(2024BAA014)

An Optimization Nesting Method for Rectangular Parts Based on Elite Genetic Strategy and Dynamic Gap-filling Mechanism

WU Dianjian1,2(), HUANG Shuaikang1(), DING Zican1, ZHANG Sanqiang1,2, CAO Weidong3   

  1. 1.Engineering Research and Design Institute of Agricultural Equipment,Hubei University of Technology,Wuhan,430068
    2.Hubei Engineering Research Center for Intellectualization of Agricultural Equipment,Wuhan,430068
    3.College Artificial Intelligence and Automation,Hohai University,Changzhou,Jiangsu,213200
  • Received:2025-11-28 Online:2026-10-25 Published:2026-11-10
  • Contact: HUANG Shuaikang

摘要:

针对多品种、不同规模的矩形件排样问题,提出了一种精英型遗传策略和空隙动态填充机制相融合的矩形件优化排样方法。根据矩形件容差ε和长宽比α两个性能参数制定矩形件排样优先级的分类规则,并采用十进制整数序列编码矩形件信息,按矩形件排样优先级生成一种精英型初始种群,通过精英型遗传策略实现算法朝着更优解的方向加速收敛。以排样过程中最低水平线为基准,结合矩形件分类规则,计算当前待排矩形件与每个空隙间的匹配度,通过搜索排样空隙找出矩形件最佳填充位置,实现空隙动态填充。通过多种算例分析ε和α对材料利用率的敏感性,以最高平均材料利用率为评价指标确定ε和α的最优值;针对大中小规模的矩形件排样算例和国内制造企业的两个真实生产案例,利用所提出方法、文献中方法和商用软件求解,对比分析其排样结果。结果表明,所提出方法在材料利用率和算法稳定性上表现出显著优势,验证了该方法的可行性和有效性。

关键词: 矩形件排样, 精英型遗传策略, 十进制编码, 矩形件分类, 空隙动态填充

Abstract:

An optimized rectangular nesting method integrating an elitist genetic strategy with a dynamic gap-filling mechanism was proposed for the rectangular nesting problem involving multiple varieties and varying production scales. The classification rules for rectangular nesting priorities were established based on two performance parameters including dimensional tolerance ε and aspect ratio α. A decimal integer sequence was used to encode part information, and an elite initial population was generated according to the defined priorities. This elite genetic strategy accelerates the convergence of the algorithm toward better solutions. Using the lowest horizontal line in the nesting process as a reference, the matching degree between the current part and each available gap was computed based on the classification rules. By searching all available gaps, the optimal placement for the part was selected, thereby achieving dynamic gap-filling. Through various case studies, the sensitivity of variable εand variable α to the material utilization rate was analyzed. Their optimal values were determined using the highest average material utilization as the criterion. Comparative experiments on large-, medium-, and small-scale nesting instances, along with two real-world production instances introduced from domestic manufacturing enterprises, were conducted using the proposed method, methods reported in the literature, and a commercial software. The proposed method demonstrates significant advantages in material utilization and algorithm stability, verifying its feasibility and effectiveness.

Key words: rectangular nesting, elite genetic strategy, decimal encoding, rectangular part classification, dynamic gap-filling

中图分类号: