摘要
针对滚动轴承振动信号特征提取在滤除干扰噪声的同时会将部分有用信号滤除,造成特征信号丢失的问题,提出了一种基于噪声辅助信号特征增强的滚动轴承早期故障诊断方法。采用广义多尺度排列熵筛选准则筛选振动信号,并通过粒子群优化算法优化Duffing振子系统参数,实现Duffing振子系统、输入信号与噪声间的最优匹配,从而提高随机共振效果,将部分背景噪声能量转移到滚动轴承早期微弱故障信号特征上,实现了早期微弱故障信号特征的增强。将所提方法应用于滚动轴承全寿命状态早期故障诊断,并与基于VMD的自适应形态学方法相比较,结果表明了该方法的有效性和可行性。
Aiming at the problem of parts of useful signals being filtered to cause loss of feature signals during filtering interference noise for the purpose of rolling bearing vibration signal feature extraction, an early fault diagnosis method for rolling bearing based on noise-assisted signal feature enhancement was proposed. The generalized multi-scale permutation entropy screening criterion was used to screen rolling bearing vibration signals, and parameters of a Duffing vibrator system were optimized using the particle swarm optimization(PSO) algorithm to realize the optimal matching among Duffing vibrator system, input signals and noise, and improve the stochastic resonance effect. Some background noise energy was transferred to rolling bearing’s early weak fault signal features to enhance them. The proposed method was applied in early fault diagnosis of rolling bearings under their life cycle state, and compared with the adaptive morphology method based on the variational mode decomposition(VMD). Results showed the effectiveness and feasibility of the proposed method are verified.
作者
周易文
陈金海
王恒
姜杰
ZHOU Yiwen;CHEN Jinhai;WANG Heng;JIANG Jie(School of Mechanical Engineering,Nantong University,Nantong 226019,China;Jiangsu Key Laboratory of 3D Printing Equipment and Application Technology,Nantong Institute of Technology,Nantong 226002,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2020年第15期66-73,共8页
Journal of Vibration and Shock
基金
国家自然科学基金(51405246)
江苏省“六大人才高峰”高层次人才(GDZB-048)
江苏省3D打印装备及应用技术重点建设实验室(南通理工学院)开放基金(2018KFKT07)
南通市3D打印技术及应用重点实验室(CP12016002)
江苏省研究生科研创新计划(KYCX17-1913)。
关键词
滚动轴承
DUFFING振子
广义多尺度排列熵
随机共振
rolling bearing
Duffing vibrator
generalized multi-scale permutation entropy
stochastic resonance
