基于BP神经元算法的电火花小孔加工穿透检测技术研究  

Research on Penetration Detection Technology of EDM Small Hole Machining Based on BP Neural Algorithm

作　　者：张海峰[1] 王佳行 梁乐意 周京华[1] 景惠静 ZHANG Haifeng;WANG Jiahang;LIANG Leyi;ZHOU Jinghua;JING Huijing(School of Electrical and Control Engineering,North China University of Technology,Beijing 100043,China)

机构地区：[1]北方工业大学电气与控制工程学院,北京100043

出　　处：《电加工与模具》2023年第3期14-19,65,共7页Electromachining & Mould

基　　金：国家重点研发计划项目(2021YFE0103800)。

摘　　要：针对电极本身产生的损耗影响电火花小孔加工的问题,提出了一种基于BP神经元分类算法的电火花小孔加工穿透检测技术。通过分析不同时期加工过程的放电频率,选择加工电压变化个数、脉冲放电的脉宽和脉间、伺服距离、加工电流作为模型特征输入量,在MATLAB软件中完成相应算法模型的建立,并综合考虑FPGA片上资源和算法耗时等因素的影响,在实验平台上实现了模型的FPGA程序移植,并通过加工实验验证了此方法的可行性。Aiming at the problem that the electrode loss affects the small hole machining,a penetration detection technology based on BP neural network classification algorithm is proposed.Through the analysis of the discharge frequency of the processing process in different periods,the number of changes in processing voltage,pulse width,pulse to pulse time,servo distance,and processing setting current are selected as the model feature inputs.The corresponding algorithm model is established in MATLAB,and the FPGA program transplantation of the model is realized on the experimental platform,taking into account the influence of factors such as FPGA on-chip resources and algorithm time consumption,The feasibility of this method is verified by machining experiments.

关 键 词：电火花加工 小孔加工 BP神经元算法 FPGA MATLAB 穿透检测 

分 类 号：TG661[金属学及工艺—金属切削加工及机床]