基于逐次变分模态分解的液压轴向柱塞泵故障特征提取方法  

Fault Feature Extraction Method of Hydraulic Axial Piston Pump Based on Successive Variational Mode Decomposition

作　　者：马景涛 汤胜楠 朱勇[3] 周涛[3] 郑智剑 MA Jingtao;TANG Shengnan;ZHU Yong;ZHOU Tao;ZHENG Zhijian(School of Mechanical Engineering,Jiangsu University,Zhenjiang,Jiangsu 212013;The State Key Laboratory of Fluid Power and Mechatronic Systems,Zhejiang University,Hangzhou,Zhejiang 310013;National Research Center of Pumps,Jiangsu University,Zhenjiang,Jiangsu 212013;Ningbo Academy of Product and Food Quality Inspection,Ningbo,Zhejiang 315048)

机构地区：[1]江苏大学机械工程学院,江苏镇江212013 [2]浙江大学流体动力基础件与机电系统全国重点实验室,浙江杭州310013 [3]江苏大学国家水泵及系统工程技术研究中心,江苏镇江212013 [4]宁波市产品食品质量检验研究院(宁波市纤维检验所),浙江宁波315048

出　　处：《液压与气动》2025年第3期100-110,共11页Chinese Hydraulics & Pneumatics

基　　金：全国重点实验室基金(GZKF-202316);宁波市青年科技创新领军人才(2023QL025)。

摘　　要：当液压轴向柱塞泵关键摩擦副出现故障时,其振动信号会出现调制现象,振动信号中的故障调制特征与特定的故障类型形成对应关系,通过信号分解可以从振动信号中提取出故障特征,进而用于故障诊断。以液压轴向柱塞泵为研究对象,利用变分模态分解和逐次变分模态分解分别对不同的仿真信号在含噪的情况下进行分解重构,综合对比了两种算法在分解性能方面的差异;最后将两种算法用于实测振动信号的故障特征提取中。结果表明:两种算法均适用于液压轴向柱塞泵的故障特征提取;逐次变分模态分解能更精确地重构出与柱塞泵故障高度相关的有效分量;变分模态分解提取到的有效分量幅值衰减更小,对微弱故障特征更加敏感。Once faults occur in the key friction pairs of hydraulic axial piston pumps,modulation phenomenon will be generated in their vibration signals.The fault modulation characteristics in the vibration signals establish a corresponding relationship with specific fault types.And fault characteristics can be extracted from the vibration signals through signal decomposition for fault diagnosis.In this paper,hydraulic axial piston pump is taken as the research object,variational mode decomposition and successive variational mode decomposition a utilized to respectively decompose and reconstruct different simulation signals under noisy conditions.The disparities in decomposition performance between the two algorithms a comprehensively compared.Eventually,the two algorithms a applied to the extraction of fault characteristics from the measured vibration signals of a hydraulic axial piston pump.The results indicate that both algorithms are suitable for extracting fault characteristics of hydraulic axial piston pump.Successive variational mode decomposition accurately reconstructs effective components highly related to the pump faults.And variational mode decomposition extracts effective component amplitudes with less attenuation,making it more sensitive to subtle fault characteristics.

关 键 词：液压轴向柱塞泵 故障诊断 故障特征提取 变分模态分解 逐次变分模态分解 

分 类 号：TH137[机械工程—机械制造及自动化]