液压泵轴承故障诊断的神经网络方法研究

The analysis of typical breakdown of electrohydraulic servo valve in DEH system for large-scale thermal power units

研究了基于集成ＢＰ网络的液压泵轴承故障诊断方法。利用频域和倒频域进行特征提取， 采用集成ＢＰ网络进行故障诊断和识别， 解决了液压泵轴承故障特征提出困难、多故障识别困难的问题。试验结果表明， 利用集成ＢＰ网络可以有效地诊断与识别液压泵轴承多故障模式， 并且具有很强的鲁棒性。

According to the structural principle of electro-hydraulic servo-valve and it's practical application in DERH system, the typical phenomena of breakdown and it's causes have been analyzed in this paper and some measures to prevent the breakdown are suggested.