摘要
重排方法通过将代表信号局部能量分布的几何中心重排到其质量中心,从而提高时频表示的聚集性和可读性,但是并没有完全消除交叉项.在此,将经验模式分解算法引入重排方法中,用于抑制重排方法在分析多分量信号时出现的交叉项.利用经验模式分解,首先将待分析的非平稳信号分解成有限个基本模式分量,它具有单分量信号的性质.对这些基本模式分量进行重排处理后的时频表示,具有良好的时频聚集性且交叉项被消除.通过对测试的柴油机爆燃阶段振动信号的重排时频分析,验证了该方法在机械故障特征提取中具有很好的应用潜力.
The reassignment method has been proved to produce a better localization of the signal components and to improve the readability of the time-frequency representation （TFR） by concentrating its energy at a center of gravity. But there are still few cross-terms. The empirical mode decomposition （EMD） is introduced into the reassignment method to suppress the interference of the cross-terms encountered in processing the multi-component signals. The multi-component signal can be decomposed into the finite intrinsic mode functions by using EMD. Each intrinsic mode function is a mono-component signal which has the characteristic of instantaneous frequency. Then, the reassignment algorithm can be applied to each of the intrinsic mode functions. Simulation analysis is presented to show the effects of this method. The measured deflagrating vibration signals of diesel engine were analyzed with the reassignment method. Experimental results indicate that this method has good potential in mechanical fault feature extraction.
出处
《测试技术学报》
2007年第1期17-22,共6页
Journal of Test and Measurement Technology
基金
河北省教育厅科研指导项目(z2004467)
唐山市机电一体化重点实验项目(04360802B-10)
唐山学院博士基金资助项目
关键词
时频分析
重排
经验模式分解
故障诊断
交叉项
time-frequency analysis
reassignment
empirical mode decomposition
fault diagnosis
cross-term
