摘要
在ＡＲＴ３神经网络模型的研究基础上，分析比较了ＡＲＴ３的分布式和压缩式识别码的模式识别情况，并且与ＡＲＴ２神经网络模式识别情况进行了比较，并将其应用在轴承故障诊断中．研究分析表明：ＡＲＴ３不仅继承了ＡＲＴ１和ＡＲＴ２神经网络的功能，而且还具备了相当强的功能和可扩展的潜力．
This paper analysis pattem identification problems of distributed recognition codes and compressive recognition codes of ART3 on the basis of ART3 neural networks modle,and compares pattem identification of ART3 with that of ART2.It is put into application on bearing diagnosis.The results of study and analysis indicate that ART3 can not only keep the functions of ART1 and ART2 neural networks but have extensible potentialities.
关键词
神经网络
谐振
模式识别
故障诊断
轴承
s:Neural Networks
Adaptive Resonance Theory :Pattem Rcognition: Failure Diagnosis.
