摘要
根据液压泵发生故障所表现出来的特征,采用小波包能量值提取的办法作为故障类型识别的特征量,采用BP神经网络对输入的特征量进行识别。实验结果表明:采用小波神经网络对液压泵故障类型的识别可以取得满意的效果。
According to feature shown during hydraulic pump fault,the sampling method of energy of wavelet packet was used to identify feature value of fault type.Then,these input feature value were identified by using BP neural network.The experimental re-sults demonstrate that satisfactory result can be obtained by using wavelet-neural network to identify fault types of hydraulic pump.
出处
《机床与液压》
北大核心
2014年第11期177-180,共4页
Machine Tool & Hydraulics
关键词
液压泵
BP神经网络
小波包
特征值
Hydraulic pump
BP neural network
Wavelet packet
Feature value
