首页 | 本学科首页   官方微博 | 高级检索  
     

基于高阶累积量的AR参数-模糊聚类法及应用
引用本文:蒋玲莉,曹宇翔,邓宗群.基于高阶累积量的AR参数-模糊聚类法及应用[J].电子测量与仪器学报,2012,26(9):812-817.
作者姓名:蒋玲莉  曹宇翔  邓宗群
作者单位:1. 湖南科技大学先进矿山装备教育部工程中心,湘潭411201;湖南科技大学机械设备健康维护省重点实验室,湘潭411201
2. 湖南科技大学机械设备健康维护省重点实验室,湘潭,411201
基金项目:国家自然科学基金(51105138、51175169)资助项目;湖南省高校科技创新团队支持计划;湖南省科技厅(2011TP4005-11)资助项目;湖南省教育厅(11C0530)资助项目
摘    要:自回归模型(autoregressive model,AR模型)是常用时序分析方法,包含了重要的系统状态信息,但一般适用于平稳过程,而真实信号多为非平稳非高斯信号,采用特别适用于非高斯信号分析的高阶累计量进行AR参数估计,以所得AR参数作为特征向量,以模糊聚类分析方法进行模式识别。将该方法应用于滚动轴承故障诊断,成功实现了滚动轴承故障类型判别与性能退化评估。

关 键 词:高阶累积量  AR模型  模糊聚类  故障诊断  状态评估

Autoregressive parameters estimation based on higher order cumulant as features combining with fuzzy cluster analysis for fault diagnosis
Jiang Lingli , Cao Yuxiang , Deng Zongqun.Autoregressive parameters estimation based on higher order cumulant as features combining with fuzzy cluster analysis for fault diagnosis[J].Journal of Electronic Measurement and Instrument,2012,26(9):812-817.
Authors:Jiang Lingli  Cao Yuxiang  Deng Zongqun
Affiliation:1.Engineering Research Center of Advanced Mining Equipment,Ministry of Education,Hunan University of Science and Technology,Xiangtan 411201,China;2.Hunan Provincial Key Laboratory of Health Maintenance for Mechanical Equipment,Hunan University of Science and Technology,Xiangtan 411201,China)
Abstract:AR model is a common time series analysis method,in which the important information about system is contained.It is generally applied to stationary signals while the real signals are non-stationary and non-Gaussian.Com-bining autoregressive(AR) model and fuzzy cluster analysis for fault diagnosis and condition assessment are proposed in this paper.The set of adjustable parameters of the AR model are estimated based on higher-order cumulant.The bearing fault diagnosis examples show that the proposed approach can be applied the condition recognition and degradation as-sessment effectively.
Keywords:higher-order cumulant  AR model  fuzzy cluster analysis  fault diagnosis  condition assessment
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号