【摘要】 针对轴向柱塞泵故障振动信号的非平稳特征,融合经验模态分解(Empirical Mode Decomposition,EMD)和实值否定选择(Real-valued Negative Selection,RNS)算法,提出了一种混合的故障诊断方法。在该方法中,首先将原始信号用EMD方法分解为若干个平稳内禀模态函数(Intrinsic Mode Function,IMF)之和,以各层能量与总能量的比值为元素构造特征向量,作为样本;然后将正常样本作为RNS算法的输入,产生检测器,对故障样本进行识别。最后用轴向柱塞泵脱靴故障样本进行诊断,正确率可达90%以上,验证了混合诊断方法的有效性。更多还原

【Abstract】 Aimed at the non-stationary characteristics of the axial plunger pump fault vibration signals,the hybrid fault diagnosis method was put forward based on the Empirical Mode Decomposition(EMD) and the Real-valued Negative Selection(RNS).In this method,the original signals are decomposed into a finite number of stationary Intrinsic Mode Functions(IMF).When the fault occurring,the energy of every IMF varies certainly and the feature vector is constructed by using the ratio of the energy of every IMF to the overall energy,regarded as the sample.And then the normal samples are used as input to the RNS algorithm for generating detectors,which can recognize the fault sample.At last,the axial plunger pump fault samples are tested using the hybrid method.The classification right rate of this method is above 90%,so it is valid to the fault diagnosis of axial plunger pump.更多还原