【摘要】 为了识别在强高斯噪声背景下振动信号的谐波成分,从三阶累积量的估计算法出发,提出了基于时间平均的三阶累积量算法,进行振动信号的谐波恢复。时间平均三阶累积量是三阶累积量的一种估计值。理论推导表明,随机相位谐波过程时间平均三阶累积量为非零值,而且其一维切片仍然是谐波过程。由此提出了一种在强高斯噪声背景下识别信号的谐波成分的频谱分析方法。该方法对抑制振动信号中的高斯噪声、正确识别其中的谐波成分十分有效,工程应用实例和信号仿真都很好地验证了该方法的正确性。更多还原

【Abstract】 For the purpose of identifying harmonic components in vibration signals with strong Gaussian colored noise,by using an estimation algorithm of time averaged three-order cumulant,a method of harmonic retrieval from the(vibration) signals is presented in the paper.Time averaged three-order cumulant is one of the estimation values of three-(order) cumulant.Theoretical derivation indicates that time averaged three-order cumulant of signals,which are from(random) phase harmonic processes,is nonzero and its one-dimension slice is also a harmonic process.Consequently,the harmonic components of the signals with strong Gaussian colored noise can be identified by the spectrum of the time(averaged) three-order cumulant.The method is very effective to restrain Gaussian noise and to identify harmonic(components) from vibration signals.Experiments and analog signals indicate that the method is feasible and right.更多还原