• CN：11-2187/TH
  • ISSN：0577-6686
机械工程学报

机械工程学报 ›› 2016, Vol. 52 ›› Issue (23): 25-33.doi: 10.3901/JME.2016.23.025

• 功能设计专栏 • 上一篇    下一篇

基于矩阵算式和蚁群算法的元功能
链设计方案优化方法*

康与云1, 唐敦兵2   

  1. 1. 临沂大学机械工程学院 临沂 276005;
    2. 南京航空航天大学机电学院 南京 210016
  • 出版日期:2016-12-05 发布日期:2016-12-05
  • 作者简介:

    康与云，男，1978年出生，博士，副教授。主要研究方向为计算机辅助创新设计，现代集成制造技术。

    E-mail：kangyuyun@lyu.edu.cn

    唐敦兵(通信作者)，男，1972年出生，教授，博士研究生导师。主要研究方向为设计理论与方法，智能制造系统。

    E-mail：d.tang@nuaa.edu.cn

  • 基金资助:
    * 国家自然科学基金(51575264)、江苏省高校青蓝工程和山东省自然科研基金(ZR2016EEM47)资助项目; 20160412收到初稿，20160903收到修改稿;

An Optimization Method for Meta-functional Chain Design Solution Based on Computational Matrix and Ant Colony Algorithm

KANG Yuyun1, TANG Dunbing2   

  1. 1. College of Mechanical Engineering， Linyi University， Linyi 276005;
    2. College of Mechanical and Electrical Engineering，Nanjing University of Aeronautics & Astronautics， Nanjing 210016
  • Online:2016-12-05 Published:2016-12-05

PDF

284

摘要：

为解决矩阵算式求解元功能链设计方案过程中缺乏优化工具的问题，提出了一种矩阵算式结合蚁群算法的优化方法。分析了相似性理论，用相似度和广义距离表征两个相邻元件的兼容性;定义了相似度矩阵，并使之与设计方案矩阵关联计算，获得了蕴含元件相似度信息的设计方案矩阵;定义了设计方案的评价模型和基于蚁群算法的优化模型，给出了评价参数、权重以及评价值的计算方法;以元件的评价得分作为信息素，以广义距离作为相邻节点路径的长度，构建了信息素矩阵和概率矩阵;将方案求解问题转化为组合优化的最优路径问题，用蚁群算法直接优化蕴含在设计方案矩阵中的方案，得到了同时满足结构需求、功能需求、评价需求的设计方案。通过某三轴伺服传送机构设计方案优化的实例，验证了方法的有效性。

关键词: 方案求解, 矩阵算式, 相似性理论, 蚁群算法, 概念设计

Abstract:

To solve the problem of the lack of optimization tools in the matrix-based solving process of the functional chain design solution, an optimization method based on computational matrix combined with ant colony algorithm is proposed. The similarity theory is analyzed, and similarity and generalized distance are used to characterize the compatibility of the two adjacent elements. The similarity matrix is defined, and the design solution matrix containing similarity information is acquired by the calculation of the similarity matrix and the design solution matrix. The design scheme evaluation model and the ant colony algorithm based optimization model are defined, and the calculation method of the evaluation criterion, weight and value is proposed. As the evaluation score of components is pheromone and as the length of generalized distance is the path of adjacent nodes, the pheromone matrix and probability matrix is structured. The problem of solution solving is converted to the optimal path problem of combinatorial optimization. The design solution contained in the design solution matrix is optimized with ant colony algorithm, and the optimized design solution meeting the functional requirements, structural requirements and the evaluation target is obtained. Finally, the effectiveness of the proposed method is verified with an example of three-axis servo drive.

Key words: ant colony algorithm, computational matrix, design solution solving, similarity theory, conceptual design

版权所有 © 2019 《机械工程学报》编辑部 | 京icp备05055788号-40