摘要
文章根据数控机床故障诊断的特点,分析了数控机床故障诊断系统的故障原因和故障源,然后引入一种带联想记忆功能的神经网络模型,并介绍了这种模型的原理和算法;最后利用所得的神经网络模型,对该故障系统的样本进行学习并记忆,再就检测或用户输入得来的故障原因进行联想回忆,得到诊断结果。实验结果表明,这种基于神经网络联想记忆的数控机床故障诊断系统简单,且能满足数控系统的故障诊断要求。
Basing on the trait of Numerical Control （NC） machine fault diagnosis, the paper analyzes the fault causes and fault sources of the NC machine fault diagnosis. Furthermore, a Nerve Net （NN） model with the associational memorial function is impoted and its theory and arithmetic is introduced. At last, with the NN model being used, the swatches of the fault diagnosis system are studied and remembered; by the fault causes being associational to remember, which are detected by sensors or input by clients, the result is got. As a result, the fault diagnosis system of the NC machine basing on NN associational memory is very simple, and the fault diagnosis both the real time and off-line are carried out.
出处
《制造业自动化》
北大核心
2008年第7期21-23,共3页
Manufacturing Automation
基金
国家自然科学基金项目(60572007)
关键词
故障诊断
神经网络
联想记忆
数控机床
fault diagnosis
nerve net
associational memory
numerical control machine
