齿轮信号特征识别的谱相关密度分析  

The Application of Spectral Correlation Density toGear Signal Character Identification

作　　者：毕果[1] 陈进[1] 何俊[1] 李富才[1] 周福昌[1] 

机构地区：[1]上海交通大学振动冲击噪声国家重点实验室,上海200240

出　　处：《上海交通大学学报》2006年第7期1084-1088,共5页Journal of Shanghai Jiaotong University

基　　金：国家自然科学基金资助项目(5017506850335030)

摘　　要：针对齿轮初期故障以调频形式表现且包络解调仅能够解调幅的现状,以一般调制信号为模型,引入解析思想,对各类型调制信号进行二阶循环平稳分析,理论推导了各自的分析结果,给出了调制信号谱相关密度分析的统一构架,证明了解析形式的谱相关密度同时提取调幅、调频特征的能力.仿真分析和齿轮故障实例验证了该分析方法对于调制特征的提取能力,表明该方法能够准确反映齿轮振动信号调制特征的变化,适用于齿轮初期故障的监测和识别.Considering the main character of the early gear fault is a frequency modulation whose envelope spectrum is unable to identify, this paper studied another demodulation method--the spectral correlation density （SCD）. Analytic principle is introduced to the SCD analysis of the general modulated signals, and the uniform theoretic frame of modulated signals is acquired after working on different kinds of modulated signals. The theory analysis verifies the advantages of the SCD to modulation signals, which can work on frequency modulation as well as amplitude modulation. The simulation and experimental results indicate that the SCD analysis is a robust method for the monitoring and character identification of early gear faults.

关 键 词：谱相关密度 齿轮 调制信号 特征识别 

分 类 号：TP206[自动化与计算机技术—检测技术与自动化装置] TP911[自动化与计算机技术—控制科学与工程]