基于神经网络的高效强力砂带磨削温度的研究  

Research on Temperature of Hi-efficient and Deep-grinding in Belt Based on Artificial Neural Networks

作　　者：符志华[1] 黄云[2] 黄智[2] 刘国华[2] 

机构地区：[1]重庆大学机械工程学院 [2]重庆大学

出　　处：《工具技术》2008年第9期9-11,共3页Tool Engineering

基　　金：国家863计划资助项目(项目编号:2006AA04Z204)

摘　　要：磨削温度高是产生磨削烧伤的主要原因,建立一个合理准确的磨削温度在线预测系统,对满足核电高压容器的高效深磨质量要求至关重要,为此建立了基于神经网络的高效深切磨削温度预测模型,并与热电偶法测得的温度进行了试验比较,发现此预测模型能达到较高的预测精度,表明了此方法有较大的实用性。To meet the quality requirements of hi-efficiency Deep-Grinding for nuclear hi-pressure vessel, it is essential to establish a reasonable and accurate prediction system for grinding temperature on line for it is the main reason for grinding burn. So, in this paper, a grinding temperature Prediction model based on artificial neural networks is made. We also make a comparison between the predicted temperature and that from the thermoeouple method and found that this prediction model to achieve higher accuracy of prediction. It also showed that this method is more practical.

关 键 词：神经网络 高效深切 磨削温度 

分 类 号：TG580.14[金属学及工艺—金属切削加工及机床] TP18[自动化与计算机技术—控制理论与控制工程]