摘要
通过选择合适的信号动态指标,分析振动信号的混合时域特征并结合图论算法分析数据特征间的相关关系,对轴承故障类型进行识别,以达到轴承故障诊断的目的。实验结果表明,无向图邻接矩阵能够从全局的角度实现对每个特征值数据的整体表达,相较于常规的基于数据分析的诊断方法,进一步提高了故障诊断的准确率。
By selecting the appropriate signal dynamic index, analyzing the mixed time domain characteristics of the vibration signal and analyzing the correlation relationship between the data features combined with the graph theory algorithm, the bearing fault types were identified, so as to achieve the purpose of bearing fault diagnosis. Experimental results show that the undirected graph adjacency matrix can realize the overall expression of each eigenvalue data from a global perspective, which further improves the accuracy of fault diagnosis compared with conventional diagnostic methods based on data analysis.
作者
王国良
任雪玉
WANG Guoliang;REN Xueyu(School of Information and Control Engineering,Liaoning Petrochemical University,Fushun 113001,China)
出处
《沈阳大学学报（自然科学版）》
CAS
2023年第1期33-41,共9页
Journal of Shenyang University：Natural Science
基金
国家自然科学基金资助项目(62073158)
辽宁省“兴辽人才”支持计划项目(XLYC1807030)
辽宁省“高校创新人才”计划项目(LR2017029)。
关键词
故障诊断
特征提取
图论
滚动轴承
数据处理
fault diagnosis
feature extraction
graph theory
rolling bearing
data processing
作者简介
王国良(1981-),男,辽宁抚顺人,教授,博士。
