中国机械工程 ›› 2015, Vol. 26 ›› Issue (15): 2096-2102.

• 智能制造 • 上一篇    下一篇

面向再设计的产品模块划分方法

周开俊1;贡智兵2;童一飞3   

  1. 1.南通职业大学,南通,226007
    2.南京康尼机电新技术有限公司,南京,210013
    3.南京理工大学,南京,210094
  • 出版日期:2015-08-10 发布日期:2015-08-04
  • 基金资助:
    国家自然科学基金资助项目(61104171);南通市科技计划资助项目(CP22013002);江苏省青蓝工程资助项目(2014) 

A Module Partition Method for Product Redesign

Zhou Kaijun1;Gong Zhibing2;Tong Yifei3   

  1. 1.Nantong Vocational University,Nantong,Jiangsu,226007
    2.Nanjing Kangni New Technology of Mechantronic Co., Ltd.,Nanjing,210013
    3.Nanjing University of Science and Technology,Nanjing,210094
  • Online:2015-08-10 Published:2015-08-04
  • Supported by:
    National Natural Science Foundation of China(No. 61104171)

摘要:

针对现有模块划分方法的不足,提出和构建了产品模块划分的统一数学模型和求解流程。在此基础上,利用零件综合模糊相似矩阵构造了基于模块聚合度和分离度的适应度函数,同时根据模块划分矩阵的特点,应用矩阵交叉遗传算法进行了产品模块划分的优化求解。最后以某天文圆顶产品为例,说明了零件综合模糊相似矩阵的获取过程,以及算法的优化求解过程。实践证明该方法快速有效。

关键词: 模块划分, 相似矩阵, 遗传算法, 矩阵交叉

Abstract:

In order to improve the defects in present methods of module partition, a unified product module partition mathematical model and solution processes were proposed. According to the characteristics of the mathematical model, a fitness function module was constructed based on the degree of polymerization and the degree of separation, then the matrix crossover genetic algorithm was applied for solving optimization module partition. Finally, taking astronomical dome products for example, the component integrated fuzzy similar matrix acquisition was illustrated, as well as the optimization algorithm for solving process. Practice has proved that this method is fast and effectives.

Key words: module partition, similar matrix, genetic algorithm, matrix-cross

中图分类号: