摘要
提出了基于EMD和AR模型的汽车变速器齿轮故障诊断方法。该方法采用EMD将齿轮振动信号 分解成若干个平稳的IMF分量,对每一个IMF分量建立AR模型,以模型的自回归参数和残差的方差作为特征向 量建立Mahalanobis距离判别函数,进而识别齿轮的工作状态。实验分析表明,该方法能有效地应用于变速器齿轮 的故障诊断。
A fault diagnosis approach for gears in automotive transmission based on empirical mode decomposition (EMD) method and auto regression (AR) model is proposed. By using EMD method, the vibration signal of gears is decomposed into a number of IMF components and then the AR model for each IMF component is established. The auto-regression parameters and the variance of remnant are taken as the eigenvectors for creating Mahalanobis distance criterion function and further identifying the gears state. Practical examples show that the proposed approach can be applied to fault diagnosis of gears in automotive transmission effectively.
出处
《汽车工程》
EI
CSCD
北大核心
2005年第1期107-110,共4页
Automotive Engineering
基金
国家自然科学基金(50275050)
高等学校博士点专项科研基金(20020532024)资助项目。
