摘要
表面粗糙度趋势分析及预测技术是计算机集成制造系统故障诊断技术发展的迫切需要。本文在讨论神经网络非线性、多因素预测原理及其拓扑结构的基础上,基于神经网络方法设计了智能型的工件表面粗糙度监测预测系统,将非线性预测和多因素预测引入表面粗糙度预测模型中,即在进行工件表面租糙度预测时兼顾了刀具磨损,从而使本系统拥有可靠和高精度的预测效果。
The study of trend analysis and forecasting method of surface roughness are an urgent need of fault diagnosis technology in CIMS. Non - linear and multi - feature forecasting theory of neural network is discussed in the paper, and is put into use in the monitoring and forecasting system of surface roughness. Tool wear state information is contained in the forecasting model too. So it is an intelligent forecasting system which has reliable and accurate forecasting result. The system will be widely used in the machinery fault diagnosis in the near future.
出处
《机床与液压》
北大核心
1996年第4期37-38,31,共3页
Machine Tool & Hydraulics
关键词
神经网络
非线性预测
因素预测
表面粗糙度
Neural network Non - linear forecasting Multi - feature forecasting Surface roughness Tool wear
