基于RBF的电火花多电源同步均衡控制算法  

Multi-Power WEDM Synchronous Equalization Discharge Control Algorithm Based on RBF

作　　者：焦银亮 罗福源[1] 杨杰 JIAO Yinliang;LUO Fuyuan;YANG Jie(College of Mechanic and Electric Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

机构地区：[1]南京航空航天大学机电学院,江苏南京210016

出　　处：《机械制造与自动化》2020年第3期170-174,共5页Machine Building & Automation

基　　金：江苏省自然科学基金优秀青年基金(BK20160084);中央高校基本科研业务费项目(NS2016056)。

摘　　要：针对"一电四线"式混合进电电火花超细密线网切割方法存在不同组间电极丝切割速率不一致、硅片切缝不均匀、切片厚度一致性差等问题,设计了一种基于RBF的电火花多电源同步均衡控制算法。在加工过程中实时调整数据模型,并结合动量增量法和梯度下降法,根据Jacobian信息和不同组间电流的差值来调整电源脉冲宽度,建立了simulink仿真模型和多电源并行放电线切割试验系统。仿真和实验结果表明,该算法收敛快,实时性强,切割质量高。The in consistency of the speed rate of slicing silicon material with hybrid power supply and the poor uniform silicon wafer and slicing thickness exist in the“one-power four-wire”super dense wire network EDM method.This paper proposes a multi-power synchronous equalization control algorithm based on RBF neural network.The Jacobian information and the difference between the currents of different groups are used to adjust the power pulse width and establish a Simulink simulation model and a multi-power parallel discharge cutting test system.The simulation and experimental results show that the algorithm above is of fast convergence,strong real-time performance,and high slicing quality.

关 键 词：半导体材料 电火花切割 混合进电 RBF神经网络 均衡放电控制 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]