【摘要】 提出了一种用于在高斯与非高斯混合有色噪声中进行谐波恢复的新方法。首先利用Hilbert变换将实数观测值变换成复数形式 ,然后应用一种特殊的高阶累计量识别非高斯噪声中的AR参数并对含噪观测值进行预滤波后 ,再应用基于统计量的方法进行谐波恢复。文中提出的方法能够用来恢复高斯与非高斯混合有色噪声中的一维实数谐波信号 ,而不用考虑谐波信号是否存在相位耦合和非高斯噪声是否对称分布 ,仿真实验可以证明更多还原

【Abstract】 A new approach to harmonic retrieval in the presence of mixed colored Gaussian and non Gaussian ARMA noises is proposed. Hilbert transformation is first used to transform the measured real value into their complex form ,then a kind of apecial high order cumulant is employed to identify the AR parameters of the non Gaussian noise.After prefiltering the observed noisy value with the identified AR parameters, cumulant based method can be applied to retrieve the harmonics. The proposed approach can be used to retrieve one dimensional real valued harmonic signals in mixed colored Gaussian and non Gaussian ARMA noises, no matter whether there is quadratic phase coupling in the harmonics and no matter whether the distribution of the non Gaussian noises is symmetrical . Simulation examples have shown its effectiveness.更多还原