极点加权模态分解及其在故障诊断中的应用

经验模态分解(EMD)是一种自适应信号分解方法,由于其能够同时提供振动信号时域和频域的局部信息,在机械故障诊断领域得到广泛应用。受EMD思想的启发,基于相邻极值加权构造均值曲线,提出一种新的自适应信号分解方法—极点加权模态分解(EPWMD)。通过仿真信号分析,将提出的EPWMD方法与EMD和局部特征尺度分解(LCD)等方法进行对比,结果表明,与EMD和LCD相比,EPWMD方法在分解性能和分解精度方面有显著提高。最后,将提出的EPWMD方法应用于转子碰摩和滚动轴承局部故障信号分析,并与EMD方法进行对比,分析结果表明,EPWMD方法不仅能够有效识别故障特征,而且其诊断效果优于EMD方法。

Extreme-pointsWeighted Mode Decomposition and Its Application to Fault Diagnosis

Empirical mode decomposition(EMD),as an adaptive signal decomposition method,has been widely applied in the mechanical fault diagnosis since it can simultaneously provide local information in the time and frequency domains of vibration signals.Inspired by EMD,a new adaptive signal decomposition method called extreme-points weighted mode decomposition(EPWMD)is proposed based on the mean curve construction of weighting with adjacent extremepoints.The proposed EPWMD method is compared with EMD and local feature scale decomposition(LCD)through simulation signal analysis.The results indicate that the EPWMD method has a significant improvement in the decomposition performance and precision compared with EMD and LCD.Finally,the proposed EPWMD is applied to the analysis of signals of the rolling bearing with local fault and the rotor system with rubbing fault.Comparing the analysis results with those of EMD,it is shown that the EPWMD can effectively identify faults and have better diagnostic results than the EMD method.