摘要
通过添加一层网络,对齿轮工作过程中的多个特征参数进行采样值的编码,使故障的特征信号转化为ART1网络可以处理的二值输入,解决了ART1网络不能处理连续信号的局限,使ART1网络能够在齿轮的故障诊断上有了进一步的应用。克服了以往诊断手段在识别有相似特征参数的不同故障上的不足。
In this thesis we have translated the characteristic signal of fault to the inputs of twovalue by adding an additional layer to code the samples of different parameters of gears in working order, so it is solved that the ART1 can＇ t be used for conducting continuous signal and the further use of ART1 in fault diagnosis in gears is realized in this thesis. The shortage of before methods in fault diagnosis for recognizing different fault with the similar characteristic parameters is resolved.
出处
《机械设计与制造》
北大核心
2006年第2期88-90,共3页
Machinery Design & Manufacture
