中文摘要：

将经验模态分解（EMD）和魏格纳-维利分布（WVD）两种方法应用于往复泵泵阀振动信号特征的提取。首先对现场采集的振动信号进行EMD分解,再对分解得到的固有模态函数（IMF）分量进行WVD分析后叠加,最终得到的信号Wigner-Ville分布可以有效抑制由于干扰项引起的频率混叠和干扰问题,有助于将原始信号在时间历程、频率成分和幅值大小三个方面的特征信息同时进行准确提取。试验结果表明,该方法能够全面、有效地表征出振动信号中所蕴含的泵阀状态信息,为后续进行泵阀状态识别和故障诊断奠定基础。

英文摘要：

The combination of empirical mode decomposition（EMD） and Wigner-Ville distribution（WVD） was applied as a tool for feature extraction from vibration data of triplex pump valves.Firstly,vibration data are decomposed into several intrinsic mode functions（IMF） with EMD,then WVD is applied to each IMF.Through the obtained WVD,the features in time,frequency and amplitude domains of the original signal can be extracted.By combining all WVDs of IMFs,the WVD-based time-frequency distributions of the original vibration signal with an effective suppression of the cross-term interference were obtained.The experimental results show that the presented method can represent the valued characteristics of vibration data,which is of great importance for condition recognition and fault diagnosis of reciprocating pump valves.