电火花小孔加工穿透检测模型仿真研究  

Simulation Research on Penetration Detection Model of Small Hole Drilling EDM

作　　者：张海峰[1] 申泽 刘建勇 王佳行 刘东辉 ZHANG Haifeng;SHEN Ze;LIU Jianyong;WANG Jiahang;LIU Donghui(School of Electrical and Control Engineering,North China University of Technology,Beijing 100043,China;Beijing Institute of Petrochemical Technology,Beijing 102600,China)

机构地区：[1]北方工业大学电气与控制工程学院,北京100043 [2]北京石油化工学院,北京102600

出　　处：《电加工与模具》2022年第2期15-20,共6页Electromachining & Mould

基　　金：国家重点研发计划项目(2021YFE0103800)。

摘　　要：针对电火花小孔加工过程中电极损耗影响加工精度和加工效率的问题,提出一种基于支持向量机分类算法的实时穿透检测仿真模型。通过对穿透发生前后加工电压和电流变化的分析,采用加工电压、加工电流、脉冲间隔设定值和低电压持续时长作为特征向量,在MATLAB软件中完成支持向量机训练模块,并在现场可编程门阵列中搭建支持向量机决策模块流程图,得出可准确、快速检测加工穿透发生的结论。Aiming at the problem that the machining electrode loss will be caused in the machining process of small hole drilling EDM,which will affect the machining accuracy and machining efficiency,a real-time penetration detection simulation model based on support vector machine(SVM)classification algorithm is proposed in this paper. By analyzing the changes of processing voltage and current before and after penetration,using processing voltage,processing current,pulse interval setting value and low voltage duration as feature vectors,the support vector machine training module is completed in MATLAB. Finally,the schematic diagram of support vector machine decision module is built in field programmable gate array(FPGA). This method can accurately and quickly detect the occurrence of machining penetration.

关 键 词：支持向量机 分类算法 穿透检测 电火花小孔加工 

分 类 号：TG661[金属学及工艺—金属切削加工及机床]