基于IUPEMD和RCMFE的往复压缩机气阀故障诊断  

Fault Diagnosis for Gas Valve of Reciprocating Compressor Based on IUPEMD and RCMFE

作　　者：宋美萍[1] 王金东[1] 赵海洋[1] 于德龙 SONG Meiping;WANG Jindong;ZHAO Haiyang;YU Delong(Heilongjiang Key Laboratory of Petroleum Mechanical Engineering,Northeast Petroleum University,Daqing Heilongjiang 163318,China)

机构地区：[1]东北石油大学黑龙江省石油机械工程重点实验室,黑龙江大庆163318

出　　处：《机床与液压》2023年第7期208-213,共6页Machine Tool & Hydraulics

基　　金：黑龙江省自然科学基金联合引导项目(LH2021E021)。

摘　　要：由于往复压缩机的振动信号非线性、非平稳性的特点,为进一步提高故障识别率,提出一种基于改进的均匀相位经验模态分解(IUPEMD)和精细复合多尺度模糊熵(RCMFE)的往复压缩机气阀故障诊断方法。采用IUPEMD方法对信号进行分解,通过不同的参数组合,利用正交性为指标选择最佳IMF分量,有效提高了IUPEMD对非平稳性信号的分解精度,减少模态混叠现象;以峭度为评价指标对分解后的IMF分量进行筛选,并重构信号,求解重构信号的RCMFE,提取故障特征向量;最后,将特征向量输入到支持向量机进行分类识别。试验结果验证了该方法的有效性和优越性。In view of non-linear and non-stationary of the vibration signal of the reciprocating compressor,in order to further improve the fault recognition rate,a fault diagnosis method for gas valve of reciprocating compressor based on improved uniform phase empirical mode decomposition(IUPEMD)and refine composite multi-scale fuzzy entropy(RCMFE)was proposed.The IUPEMD method was used to decompose the signal,through different parameter combinations,orthogonality was used as the index to select the best IMF component,which could effectively improve the accuracy of IUPEMD’s decomposition of non-stationary signals and reduce modal aliasing.Using kurtosis as the evaluation index to filter the decomposed IMF components and reconstruct the signal,the RCMFE of the reconstructed signal was solved,and the fault feature vector was extractd.Finally,the feature vectors were input to the support vector machine for classification and recognition.The test results verifies the effectiveness and superiority of this method.

关 键 词：改进的均匀相位经验模态分解 精细复合多尺度模糊熵 气阀 故障诊断 

分 类 号：TH165.3[机械工程—机械制造及自动化]