基于紧致型小波神经网络的往复泵故障诊断

为了对往复泵的故障进行正确诊断，提出基于紧致型小波神经网络的往复泵故障诊断方法。以往复泵单个泵缸内的压力信号作为系统特征信号通过小波包分解来提取故障特征向量，同时将此特征向量作为小波神经网络的输入，利用小波神经网络对故障做进一步的精确实时诊断。通过对往复泵液力端多故障诊断实例的检验表明，该系统故障诊断正确率达到94 %以上。

Fault Diagnosis Technology of Reciprocating Pumps Based on Compact Wavelet Neural Network

A method of the reciprocating pump fault diagnosis is proposed based on inlay model Wavelet Neural Network, in order to determine the fault type of the reciprocating pump accurately. This paper using the reciprocating pump single cylinder pressure as the characteristic signal of the system, extracting feature vector of the faults by wavelet packet decomposition, and at the same time, using this feature vector as the input signal of the wavelet neural network to determine the type of the fault. The examples of faults diagnosis of fluid end on a reciprocating pump show that the system fault diagnosis accuracy reached more than 94%.