摘要
由于数控机床方案设计实例的多样性和复杂性,实例检索和修改过程中很难获取相关知识。针对此情况,结合数控机床方案多层次结构特点,运用相似度的高低和遗传算法对实例进行检索与修改。创建数控机床多层次结构实例库,通过计算实例相似度,检索出实例库中与目标实例相似度较高的实例;利用遗传算法对实例进行修改,将检索出的相似实例作为遗传算法的初始种群,通过交叉、变异选择出优秀个体参与遗传操作;以数控机床床身为例,通过编写程序求取遗传算法的最优设计方案。结果表明:运用此方法可快速找到与设计问题相似度较高的方案。
Due to the diversity and complexity of the design example of CNC lathe,it is difficult to obtain the relevant knowledge in the process of retrieval and modification.In this case,combined with the characteristics of the multi layer structure of CNC machine tools,similarity and genetic algorithm were used to retrieve and modify the case.A multi level structure of the CNC lathe example library was created;through the calculation of the similarity of examples,the example of a high degree of similarity with the target instance was retrieved.Then the genetic algorithm was used to modify the instance,and the similar instances were retrieved as the initial population of genetic algorithm.Through crossover and mutation,the excellent individuals were selected to participate in the genetic operation.Finally,the CNC lathe bed was taken as an example,through programming the optimal design scheme for genetic algorithm was gotten.The result shows that this method can be used to quickly find a high degree of similarity with the design problem.
作者
栗江
张慧
LI Jiang;ZHANG Hui(Department of Mechanical and Electrical Engineering,Guangzhou Nanyang Polytechnic,Guangzhou Guangdong 510925,China;School of Industrial Engineering,Ningxia Vocational and Technical College,Yinchuan Ningxia 750021,China)
出处
《机床与液压》
北大核心
2018年第19期93-96,共4页
Machine Tool & Hydraulics
基金
广东省科技厅项目(2016020A0101082)
关键词
遗传算法
数控机床
实例检索
实例修改
相似度
Genetic algorithm
CNC lathe
Case retrieval
Case modification
Similarity
