基于峰值电流与脉冲宽度的微细电极控形研究

Research on the shape wear control of micro electrode based on peak current and pulse width

针对微细电火花孔加工时因微细电极形状损耗难以控制导致的微孔加工精度不高问题,提出利用峰值电流和脉冲宽度两个重要加工参数控制微细电极形状损耗的方法,并运用差分进化(Differental Evolution,DE)算法优化的支持向量机(Support Vector Machine,SVM)(DE-SVM)方法建立了微细电极形状损耗的分类预测模型。研究表明:该方法是可行的,对于给定的试验数据,相比常用的粒子群(Particle Swarm Optimization,PSO)算法优化的SVM(PSO-SVM)方法和遗传算法(Genetic Algorithm,GA)优化的SVM(GA-SVM)方法,DE-SVM方法能够获得分类准确率高且拟合度合理的分类预测模型;不同微细电极形状损耗形式具有紧密的相关性,在较小的峰值电流(4~20 A)和较大的脉冲宽度(>5μs)条件下易获得底部规整的微孔。研究成果从微细电极形状损耗控制角度出发,为提高微细电火花孔的加工精度提供了一种思路。

In order to solve the problem of low precision in micro-EDM drilling due to the difficulty in controlling the shape wear of micro-electrode,a method was put forward,which using the peak current and pulse width to control the shape wear of electrode,and the shape wear of electrode classification prediction model was established by using Differental Evolution algorithm optimized Support Vector Machine(DE-SVM).The research shows that this method is feasible.For given experimental data,compared with the Particle Swarm Optimization optimized SVM(PSO-SVM)and the Genetic Algorithm optimized SVM(GA-SVM),DE-SVM method can obtain a classification prediction model with high classification accuracy and reasonable fitting degree.The wear pattern of different electrode shapes wear has a close correlation.Under the condition of small peak current(4~20 A)and large pulse width(>5μs),it′s easy to get the micro holes,which the bottom is regular.The results of research provide a way to improve the machining precision of micro EDM drilling from the shape wear control of micro electrode.