摘要
针对滚动轴承复合故障特征存在交叉影响又受环境噪声干扰、分离诊断困难问题,提出一种基于改进奇异谱分解(SSD)和多点最优最小熵解卷积调整(MOMEDA)的滚动轴承复合故障分离诊断方法。首先,为了克服SSD分解层数需要凭经验设定而难以选到最优分量的缺点,提出相关峭度图方式优选复合故障中不同故障各自最优分解层数的改进SSD方法;其次,对选出的各最优分量,利用MOMEDA能够降噪、增强冲击特征的优点,进一步削弱其中残存的交叉及干扰成分;最后,由包络分析诊断出复合故障。实验信号分析结果表明:改进SSD方法能准确确定复合故障中不同故障的各自最优分解层数,经MOMEDA处理后的故障特征更明确、诊断更可靠,实现了轴承复合故障的有效分离和故障类型的准确诊断,为轴承复合故障的分离诊断提供了一条途径。
Cross effect and environment noise interference exist in rolling bearing compound fault.In order to solve the separation problem of fault characteristics,a rolling bearing compound fault separation diagnosis method based on improved singular spectrum decomposition(SSD)and multipoint optimal minimum entropy deconvolution adjusted(MOMEDA)was proposed.Firstly,in order to overcome the disadvantage that the number of decomposition layers of SSD needed to be set by experience and it was difficult to select the optimal component,an improved SSD method based on correlation kurtosis graph was proposed to optimize the respective decomposition layers of different faults in composite faults.Then,the remaining cross and interference components in the selected optimal components were further weakened by MOMEDA that had the ability of reducing noise and enhancing the impact characteristics.Finally,the compound fault was diagnosed by envelope analysis.Experimental signal analysis results show that the improved SSD method can accurately determine the optimal decomposition layers of different respective faults in the composite fault,and the fault characteristics processed by MOMEDA are more obvious and the diagnosis is more reliable.The effective separation of bearing compound fault and accurate diagnosis of fault types are relized.It provides a way for the separation diagnosis of bearing compound fault.
作者
刘尚坤
张伟
范壮壮
孔德刚
张秀花
LIU Shang-kun;ZHANG Wei;FAN Zhuang-zhuang;KONG De-gang;ZHANG Xiu-hua(School of Mechanical and Electrical Engineering,Hebei Agricultural University,Baoding 071001,China)
出处
《组合机床与自动化加工技术》
北大核心
2022年第6期138-141,145,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
河北省省属高校基本科研业务费项目(KY202014,QN2021224)
保定市科技计划项目(2172P003)
河北农业大学引进人才科研专项(YJ201814)
河北省引进留学人员资助项目(C20200336)。
关键词
改进奇异谱分解
相关峭度
多点最优最小熵解卷积调整
滚动轴承
复合故障
improved singular spectrum decomposition optimization
correlation kurtosis
multipoint optimal minimum entropy deconvolution adjusted
rolling bearing
compound fault
