摘要
介绍了双向联想记忆(BAM)学习算法神经网络和误差后向传播(BP)神经网络在数控系统故障诊断中的应用实例。这两种神经网络在数控系统故障诊断方面都有着重要的应用价值,都能实现故障的联想、容错、自学习等功能。但因为各自具有的特点不同,又使得BP网络更适于故障样本多的情况。
:This paper presents examples of application of bidirectional associative memory(BAM) and back propagation(BP) to CNC fault diagnosis. Both of the two methods are of great application value and have the functions including fault association, allowing error and studying by themselves. But because they have some different characteristics, BP apply to fault diagnosis system which have lots of training patterns, and BAM apply to fault diagnosis system asking to correct the false detected signals.
出处
《航空制造工程》
1997年第3期22-24,共3页
Aviation Engineering & Maintenance
基金
航空科学基金