CAJ | 학술논문

针对经验模态分解（EMD）的模态混淆问题，提出了一种新的抑制模态混淆的方法---基于伪极值点假设的经验模态分解（PEEMD）。与总体平均经验模态分解（EEMD）通过添加白噪声再进行总体平均的方式不同，PEEM D通过定义最小极值尺度，并用其度量其他极值尺度，通过增加伪极值点的方式来均匀化尺度，有效地抑制了模态混淆的产生。详细介绍了PEEM D方法，并通过仿真信号将其与EM D和EEM D进行了对比，最后，将PEEM D应用于转子碰摩故障的诊断中。仿真和实测信号结果表明，PEEM D在分量的精确性和抑制模态混淆的产生等方面要优于EM D和EEM D ，是一种有效的信号分解方法。
침대경험모태분해（EMD）적모태혼효문제，제출료일충신적억제모태혼효적방법---기우위겁치점가설적경험모태분해（PEEMD）。여총체평균경험모태분해（EEMD）통과첨가백조성재진행총체평균적방식불동，PEEM D통과정의최소겁치척도，병용기도량기타겁치척도，통과증가위겁치점적방식래균균화척도，유효지억제료모태혼효적산생。상세개소료PEEM D방법，병통과방진신호장기여EM D화EEM D진행료대비，최후，장PEEM D응용우전자팽마고장적진단중。방진화실측신호결과표명，PEEM D재분량적정학성화억제모태혼효적산생등방면요우우EM D화EEM D ，시일충유효적신호분해방법。
A novel method PEEMD was proposed for restraining the mode mixing problem of EMD .Compared with ensemble EMD(EEMD) ,which was by adding white noise before ensemble and averaging ,PEEMD fulfilled this aim by defining the least extrema scale(LES) to measure other extre-ma scales and by adding pseudo-extrema to homogenize the scale distribution .EMD and PEEMD are introduced herein firstly and then PEEMD was compared with EMD and EEMD by analyzing simula-tion signals .Also ,PEEMD was applied to diagnose rotor rubbing faults ,and the results indicate that PEEMD is more accurate in components and more efficient in inhibition of mode mixing than that of EM D and EEM D .