【摘要】 为更准确提取旋转机械故障特征,提出了基于全矢二元经验模态分解(Bivariate Empirical Mode Decomposition,BEMD)的故障特征提取方法。该方法首先通过多传感器正交采集旋转机械故障同一截面上的振动信号,并将其组成一个复数;然后运用BEMD将复数按旋转速度从高到低的顺序自适应地分解到各自的频带,得到系列复固有模态分量(Complex Intrinsic Mode Functions, CIMFs);提出复数相关系数的概念,并用于组合CIMFs得到新的复旋转分量以防同一频率的信号被分解到不同的CIMFs;最后,运用全矢谱融合组合后的CIMFs的特征信息,得到幅频、角度和进动方向等信息。与全频谱方法的对比试验结果表明该方法的有效性。更多还原

【Abstract】 A method of full vector bivariate empirical mode decomposition(BEMD) was proposed to more accurately extract fault characteristics of rotating machinery. Firstly, vibration signals at fault position’s cross-section of rotating machinery were collected with orthogonally located multi-sensor to form a complex. Then, BEMD method was applied to adaptively decompose the complex into different frequency bands according to rotating speed values’ high to low turn to obtain complex intrinsic mode functions(CIMFs). Finally, the full vector spectrum technique was used to fuse characteristic information of CIMFs to acquire information of amplitude-frequency, phase-frequency and precession directions. The test results using the proposed method were compared with those using the full frequency spectrum one. It was shown that the proposed method is effective.更多还原