基于AR高阶谱及其关联维数分析的减压阀故障诊断  

Fault Diagnosis of Reducing Valve Based on AR Higher-Order Spectra and Their Correlation Dimension Analysis

作　　者：陈庆堂[1] 黄宜坚[2] 

机构地区：[1]莆田学院机电工程学院,福建莆田351254 [2]华侨大学机电及自动化学院,福建厦门361021

出　　处：《机械设计与研究》2014年第4期110-115,共6页Machine Design And Research

基　　金：国家自然科学基金课题(50975098);福建省教育厅中青年教师科技A类资助(JA13434)

摘　　要：为了识别减压阀的工作状态,建立了减压阀不同工作状态采集信号的时间序列AR模型,绘制了AR三谱、双谱及其切片谱图,计算了各切片谱的关联维数,综合分析了不同工作状态系统的谱图及关联维数变化。分析结果表明,AR三谱、双谱、各切片谱及其关联维数各自对工作状态变化的敏感性不同,双谱对角切片的关联维数、三谱及其切片谱对工作状态变化较敏感,更适合用于减压阀故障诊断,同时AR三谱、双谱的一维切片谱及其关联维数在反映系统动力学特性方面分别存在对应关系,可以将切片谱和关联维数相结合对减压阀进行故障诊断分析。In order to identify the working states of reducing valve, the time series AR model of the vibration signal in different states for reducing valve is founded, the bispectra, trispectra and their slices are plotted, correlation dimensions of the spectra slices are calculated, the change of the spectra and their correlation dimentions are analyzed under normal and fault working states. The result shows that the higher-order spectra, their slice spectra and correlation dimentions of slice spectra each shows different sensitivity to the change of working states, correlation dimentions of bispectra~ slice, trispectra and its slice spectra are more sensitive than others, so they are more suitable for fault diagnosis of reducing valve; the correspondence exists between each one-dimension slice spectrum and correlation dimension in reflecting dynamic characteristics of system, fault diagnosis and analysis can be done for reducing valve with the combination of slice spectrum and correlation dimension.

关 键 词：高阶谱 关联维数 故障诊断 

分 类 号：TB137.522[理学—物理]