基于随机模糊神经网络的刀具磨损量软测量技术  

TOOL WEAR ESTIMATION BY SOFT-SENSING TECHNOLOGY BASED ON STOCHASTIC FUZZY NEURAL NETWORK

作　　者：王军平[1] 敬忠良[2] 王安[1] 

机构地区：[1]西北工业大学自动控制系,西安710072 [2]上海交通大学航空航天信息与控制研究所,上海200030

出　　处：《信息与控制》2002年第6期534-537,共4页Information and Control

基　　金：国家教育部留学回国人员基金;跨世纪优秀人才培养计划基金;航天科技创新基金资助

摘　　要：刀具磨损检测对于提高加工过程的自动化、高精度化、智能化具有重要意义 .本文通过检测电流信号基于随机模糊神经网络建立了刀具磨损量的软测量模型 .该模型的创新之处在于利用切削参数实时地调整网络的部分参数 ,从而可以减小切削参数与电流信号之间关系对于刀具磨损估计的影响并且使得模型具有动态性、实时性 .实验验证表明该方法是正确而有效的 .Tool wear measurement would be a great significance for improving the automation, accuracy and intellegence of the manufacturing process. Through measuring the electric current signal, the soft sensing model used for tool wear estimation based on stochastic fuzzy neural network(SFNN) is presented in this paper. In the model, the cutting parameters are used to adjust several parameters of SFNN on line, so the influence on the tool wear estimation by the relation of the electric current signal and the cutting parameters is eliminated and the model is dynamic. The experimental results have shown the effectiveness of this method.

关 键 词：随机模糊神经网络 刀具 磨损量 软测量技术 数控系统 自动化加工 

分 类 号：TG71[金属学及工艺—刀具与模具]