摘要
针对故障行星齿轮箱的扭振信号具有非稳定、非线性的特点,提出基于H-P滤波与变分模态分解(Variational Mode Decomposition,VMD)相结合的行星齿轮故障时域特征分离方法。H-P滤波基于竞争性趋势估计算法,能保存扭振信号中的分段线性特征。滤波后的信号内容丰富,VMD能将这些内容分解在不同的模态分量中,故障冲击特征突出。通过分析不同负载下的扭振信号,证实该方法能够提取周期脉冲,诊断出行星齿轮故障。对于实际工程应用中的复杂信号,该方法提供了有效解决途径,具有一定的工程应用价值。
The torsional vibration signal of the faulty planetary gearbox has the characteristics of non-steady and non-linearity,time-domain fault characteristics are difficult to extract. In order to solve this problem,a time-domain feature separation method based on H-P filtering and VMD is proposed. H-P filtering is based on a competitive trend estimation algorithm,it can preserve the piecewise linear features in torsional vibration signals. The filtered signal is rich in content,VMD can decompose these contents into different modal components,fault impact characteristics prominent. By analyzing the torsional vibration signals under different loads,confirmed that the method can extract periodic pulses,diagnosis of planetary gear failure. Actual engineering applications are mostly complex signals,not conducive to diagnostic research,the method provides an effective solution,has a certain value of engineering applications.
出处
《失效分析与预防》
2018年第1期29-33,共5页
Failure Analysis and Prevention
基金
国家自然科学基金(51465040)
