HDI板孔群的K-means聚类及加工路径优化研究  

Research on K-means Clustering and Path Optimization Process of HDI Plate Group

作　　者：张明禄 李忠学[1] 陈辅斌 ZHANG Ming-lu;LI Zhong-xue;CHEN Fu-bin(School of Mechatronic Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)

机构地区：[1]兰州交通大学机电工程学院,兰州730070

出　　处：《组合机床与自动化加工技术》2018年第7期134-137,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金资助项目(71361019);兰州交通大学青年基金(2015007)

摘　　要：电路板的激光打孔路径优化是一个NP完全问题,随着孔群数量的增加,其复杂性和解空间会成指数性增长。文章采用K-means算法和改进模拟退火算法,并结合孔群加工的数学模型研究HDI激光打孔的路径优化。首先采用K-means算法聚类分析得到相似度高的孔群,初始种群运用贪婪选择策略,降低起始搜索范围,提高收敛效率;寻优过程中加入遗传算法的DIM(Displaced Inversion Mutation)变异算子,避免模拟退火算法陷入局部最优解。仿真试验结果表明,该方案求解出的最优路径比采用模拟退火算法有明显改善。The optimization of laser drilling path of circuit board is a NP complete problem.With the increase of microviasits complexity and solution space will increase exponentially.In this paper,the K-means algorithm and the improved simulated annealing algorithm are combined with the mathematical model of hole processing to study the optimization of HDI laser drilling.Firstly,it obtains high Similar microvias by K-means clustering algorithm;In the initial population,the greedy selection strategy is adopted to reduce the initial search search range and improve the convergence efficiency;The genetic algorithm is added in the searching process of DIM(Displaced Inversion Mutation)mutation operator to avoid local optima.The simulation results show that the optimal path solved by the proposed scheme is better than the simulated annealing algorithm.

关 键 词：激光打孔 K-MEANS聚类算法 改进模拟退火算法 路径优化 

分 类 号：TH16[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]