基于GA-GRNN的复合铣床立柱优化  

Column Optimization of Compound Milling Machine Tool Based on GA-GRNN

作　　者：高亚洲[1] 史耀耀[1] GAO Ya-zhou SHI Yao-yao(The key Laboratory of Contemporary Design and integrated Manufacturing Technology Ministry of Education China, Northwestern Polytechnical University, Xi＇ an 710072, China)

机构地区：[1]西北工业大学现代设计与集成制造技术教育部重点实验室,西安710072

出　　处：《组合机床与自动化加工技术》2016年第10期17-20,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家科技重大专项资助项目(2013ZX04001-081)

摘　　要：立柱是复合铣床的主要承力部件,其质量直接影响机床的刚性和动态性能,进而影响加工质量,因此对有必要对复合铣床立柱质量进行优化设计。首先采用灵敏度分析法,获得影响立柱质量的敏感尺寸参数;其次基于均匀试验和GRNN神经网络分别建立立柱质量、前2阶固有频率和最大变形量的模型,并通过遗传算法对模型方程进行寻优求解,得出尺寸参数最优解组合;最后在优化后立柱的最大变形量不超过原立柱最大变形量的情况下,优化后的机床立柱的质量减轻了10.09%,前2阶频率分别提高了3.10%、2.42%,证明GA-GRNN优化机床立柱是可靠有效的,可以将其推广到更广泛的领域。The column is the main bearing components of compound milling machine tool,and the rigidity and dynamic performance of machine tool is affected by its＇ mass directly,and then so is the processing quality,so it is necessary to optimize and design the mass of compound milling machine tool. Firstly,the sensitivity analysis is used to obtain the sensitive dimension parameters that have influence on the column mass.Secondly,the approximate models of the mass of the column,the first two order frequency and the maximum deformation are established based on the uniform experiment and GRNN. And the approximate models are optimized by genetic algorithm to obtain the optimum combination of the dimension parameters. Finally,with the static and dynamic performance of the original column structure being almost the same,the mass of the optimized machine tool column is reduced by 10. 09% and and the first two order frequency are increased by 3. 10% and 2. 42% respectively,which proves GA-GRNN is reliable and effective to optimize the machine tool column and it can be extended to more wide domain.

关 键 词：遗传算法 广义回归神经网络 均匀实验 灵敏度分析 

分 类 号：TH162[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]