斯特林发动机振动信号分析及谱估计

Stirling Engine Vibration Signal Analysis and Spectrum Estimation

为了检测斯特林发动机运行状态,针对斯特林发动机在运行过程中振动信号产生机理,采用了经验模态分解与自回归模型相结合的方法对振动信号进行分析,设计了振动检测系统.通过选取故障信息的本征模函数进行功率谱估计,提取滚动轴承故障特征.测试结果表明:经验模态分解可自适应地分解非平稳信号,生成的本征模函数可提取信号内在的本质特征.对自回归模型进行功率谱估计,提取振动状态异常信号.经实验验证,故障情况与真实异常状况吻合,可有效检测斯特林发动机运行过程中的故障特征.

In order to detect the operating state of stirling engine and evaluate the service life ,the mechanism of vibration signals based on stirling engine running is studied ,the vibration signals are analized using empirical mode decomposition and autoregressive model ,and the vibration detection systemn is designed .The intrinsic mode function of fault information is solved for power spectrum estimation to extract fault feature of rolling bearing .Then performance results show ：The non-stationary signal is decomposed by empirical mode decomposition adaptively and the inherent nature signal is extracted from intrinsic mode function .The autoregressive model is estimated in power spectrum to extract the abnormal vibration signals .After experimental verification ,the fault condition has been testified accord with true abnormal condition ,w hich effectively detects the stirling engine running state .