摘要
研究了基于小波包分解-局部均值分解算法(WPD-LMD)和多尺度熵(MSE)的滚动轴承故障诊断方法.通过传感器获取滚动轴承故障信号,建立多维信号关联矩阵区分噪声信号与故障信号,以扩展动模式分解(EDMD)方法升维观察信号,预估滚动轴承故障源信号数量.设定源信号筛选的相关程度指标,以奇异值分解定理表征滚动轴承故障信号,基于WPD-LMD分解信号频段,获取临界阈值优选故障信号特征.采用结构化理论处理高维信号,设定相类似信号共享权值,以非线性规则函数增强信号特征,划分滚动轴承故障源信号类型.基于MSE对应故障信号模态分量,围绕频率中心构建约束分量模型,对应信号频谱空间诊断滚动轴承故障类型.结果表明:该方法可以完成99.5%的去噪效果,对不同类型故障问题的诊断识别率最高为99.4%,具有较好的应用效果.
A rolling bearing fault diagnosis method based on wavelet packet decomposition-local mean decomposition algorithm(WPD-LMD)and multi-scale entropy(MSE)is studied.The rolling bearing fault signal is obtained by the sensor,and the multi-dimensional signal correlation matrix is established to distinguish the noise signal from the fault signal.The extended dynamic mode decomposition(EDMD)method is used to increase the dimension of the observed signal and estimate the number of rolling bearing fault source signals.The correlation degree index of source signal screening is set,the rolling bearing fault signal is characterized by singular value decomposition theorem,and the signal frequency band is decomposed based on WPD-LMD to obtain the critical threshold to optimize the fault signal characteristics.The structured theory is used to process high-dimensional signals,set the shared weights of similar signals,enhance the signal characteristics with nonlinear regular functions,and divide the types of rolling bearing fault source signals.Based on the modal component of MSE corresponding fault signal,the constrained component model is built around the frequency center,and the fault type of rolling bearing is diagnosed in the corresponding signal spectrum space.The results show that this method can achieve 99.5%denoising effect,and the highest diagnostic recognition rate for different types of fault problems is 99.4%,which has application effect.
作者
王琳琳
WANG Linlin(Department of Mechanical Engineering,Anhui Vocational College of Metallurgy Science and Technology,Maanshan 243000,Anhui,China)
出处
《韶关学院学报》
2023年第6期24-29,共6页
Journal of Shaoguan University
基金
安徽省教育厅自然科学研究项目“基于MATLAB的智能PID控制器的研究”(KJ2020A1151)。
