Study on Fault Diagnosis Based on Wavelet Neural Network for Hydraulic Pump

摘要：

Wavelet analysis can plot out the signal to different frequency channels, and it has the ability of characterizing the local feature of the signal at both time domain and frequency domain. Analyzing and processing the signal with Wavelet transform overcomes the limitation of the conventional Fourier transform, its credible and effective. The features of hydraulic pump fault information are extracted by wavelet analysis and entered into the neural networks as the input characteristic vectors. Diagnose the boot pull-off fault of the pump using the relax-model which is the combination of wavelet analysis and neural network. The testing result shows that the method can be applied in hydraulic pump fault diagnosis efficiently.

关键词： wavelet analysis neural network fault diagnosis

作者: Jiang Wanlu Lu Jingli

作者单位: Heavy Machinery Fluid power Transmission and Control Laboratory, Yanshan University, Qinhuangdao, Hebei, 066004, China

会议类型: 国际会议

会议名称: The Fifth International Symposium on Fluid Power Transmission and Control(ISFP2007)（2007年国际流体动力传输与控制学术会议）

会议地点: 北戴河

会议语种:英文

页码: 1085-1089

在线出版日期: 2007-06-06（万方平台首次上网日期，不代表论文的发表时间）