机械工程

基于遗传算法的模具电极调度问题求解

展开
  • 1.华南理工大学 机械与汽车工程学院,广东 广州 510640; 2.珠海格力精密模具有限公司,广东 珠海 519070; 3.广州民航职业技术学院 飞机维修工程学院,广东 广州 510403
夏琴香(1964-),女,博士,教授,主要从事先进成形与模具及装备技术研究。

收稿日期: 2021-05-10

  修回日期: 2021-08-03

  网络出版日期: 2021-08-11

基金资助

广东省重点领域研发计划项目(2019B090918004)

Die Electrode Scheduling Problem Solution Based on Genetic Algorithm

Expand
  • 1. School of Mechanical and Automotive Engineering, South China University of Technology, Guangzhou 510640, Guangdong, China; 2. Zhuhai Gree Precision Mold Company, Zhuhai 519070, Guangdong, China;  3. Aircraft Maintenance Engineering College, Guangzhou Civil Aviation College, Guangzhou 510403, Guangdong, China
夏琴香(1964-),女,博士,教授,主要从事先进成形与模具及装备技术研究。

Received date: 2021-05-10

  Revised date: 2021-08-03

  Online published: 2021-08-11

Supported by

Supported by the Key-Area Research and Development Program of Guangdong Province(2019B090918004)

摘要

针对模具电极数控加工和电火花加工两个阶段的调度问题,建立了具有批处理和相关性特征的数学模型。以模具零件拖期量最小为目标,将模具电极调度问题的求解分为批处理和批调度两个层次,第1层提出根据相关性原则解决批处理问题,并设计出相关性优先分批算法;第2层提出采用遗传算法解决批调度问题,并提出一种基于动物配种的策略对传统的遗传算法进行了改进。基于Matlab软件,开发出模具电极调度程序,实现了上述两个层次的求解过程,并在24种仿真算例下进行了实验测试。结果表明,所设计的模具电极调度算法能有效的解决模具电极批处理和批调度问题;所提出的配种策略能显著提高传统遗传算法的求解质量,使样本内的模具零件拖期量最多减少16.71%。

本文引用格式

夏琴香, 李凯, 马骏, 等 . 基于遗传算法的模具电极调度问题求解[J]. 华南理工大学学报(自然科学版), 2022 , 50(3) : 80 -87 . DOI: 10.12141/j.issn.1000-565X.210284

Abstract

In view of the scheduling problem of die electrode in CNC and EDM stages, a mathematical model with batch processing and correlation characteristics was established. To minimize the tardiness of die parts, the solving process of die electrode scheduling problem was divided into two stages: batch processing and batch scheduling. In the first stage, the batch processing problem was solved according to the principle of correlation, and the correlation priority batch algorithm was designed. In the second stage, the batch scheduling problem was solved by genetic algorithm, and a strategy based on animal breeding was proposed to improve the traditional genetic algorithm. The die electrode scheduling program was developed based on MATLAB software to realize the above two stages of solving process, and the test was carried out under 24 kinds of simulation examples. The results show that, the designed die electrode scheduling algorithm is effective for solving the batch processing and batch scheduling problem of die electrode; the proposed strategy of breeding can significantly improve the quality of traditional genetic algorithm solution; and the tardiness of die parts in the examples can be reduced by 16.71% at most.
文章导航

/