基于奇异值分解技术的轮边减速器故障信号降噪研究

利用改进的奇异值分解技术,用仿真信号验证了该技术对轮边减速器齿轮故障特征提取的有效性,并从模拟信号中提取出了故障特征频率。研究发现,噪声的奇异值分布趋于直线,凸显出了有用信号的奇异值,有利于特征信号提取。仿真结果表明,该方法能在强噪声背景下提取行星系统齿轮故障特征,为轮边减速器故障诊断提供了一个新的思路。

Singular-Value-Decomposition-Based Noise-Reducing Study of the Fault Signals of the Reduction Gear

By means of the singular value decomposition technology,the simulated signals are used to verify the reliability of the technique in extracting the characteristic frequencies of the fault from the simulated signals.The study shows that the singular value distribution of the noise tends to be linear,and the singular values of the useful signal are easily extracted.The results of the simulation show that the improved method is effective in detecting the fault signals of the planetary gearbox in the background of strong noise and provides a new thinking way for the fault diagnosis of reduction gears.