面向电加工的专用数控系统中间隙状态的智能控制  

Intelligent Control of Gap State in Computer Numerical Control System for Electrical Discharge Machining

作　　者：陈智殷[1,2] 于东[2] 胡毅[2,3] 彭鹏[1,2] 

机构地区：[1]中国科学院研究生院,北京100039 [2]中国科学院沈阳计算技术研究所高档数控国家工程中心,沈阳110171 [3]沈阳高精数控技术有限公司,沈阳110171

出　　处：《小型微型计算机系统》2014年第12期2775-2779,共5页Journal of Chinese Computer Systems

基　　金："高档数控机床与基础制造装备"国家科技重大专项(2013ZX04007-011)资助

摘　　要：为了提高对电加工间隙状态识别的准确度、测控的自适应性和加工的稳定性,解决传统间隙状态识别逻辑电路复杂等问题,在总结目前间隙状态测控方法的基础上,结合模糊控制和BP神经网络的方法实现电加工间隙状态的智能控制.模糊控制输出加工间隙相对值作为速度比率,结合插补器最大速度,控制伺服进给或后退.BP神经网络通过在线分析加工效率,判断模糊控制效果,训练样本,实时改变电压阈值,间接调整模糊推理规则,改进模糊控制性能,提高自适应性.通过仿真数据对比传统间隙识别方法,验证用C++编写的智能控制程序对间隙状态识别的性能.In order to improve the accuracy of the EDM gap state recognition and measurement, the adaptability and stability of the control process and the complexity of the traditional gap state identification logic, summarizes the current gap state monitoring and con- trolling method, achieve the intelligent control of the machining gap state based on fuzzy control and BP neural network. Fuzzy control output machining gap relative value as the speed ratio, combined with the maximum speed of the interpolator, to control servo feed or backward. By online analytical processing efficiency, BP neural network deduces the fuzzy control effect, train samples, and change the voltage threshold, indirect adjust fuzzy inference rules to improve the performance of fuzzy control, improve self-adaptability. With processing simulation data and compared with the traditional gap state identification method, verify the performance of intelligent con- trol program written by C ＋ ＋ in gap state recognition.

关 键 词：电加工间隙状态 数控系统 智能控制 BP神经网络 模糊控制 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]