基于高阶累积量Burg自适应算法的间谐波估计

Inter-harmonic Estimation Using Burg Adaptive Algorithm Based on Higher-order Cumulant

Burg算法适合处理间谐波信号,但是在有噪声的环境下,会产生错误的谱估计结果。采用一种基于高阶累积量的Burg自适应算法估计间谐波信号,该方法不受高斯噪声的影响。首先,利用高阶累积量对高斯噪声不敏感的特性,改进Burg算法的预测误差标准。其次,为了降低引入高阶累积量而增加的计算量,反射系数的求取采用递推形式。最后,利用谐波信号的4阶累积量对角切片求取各成分的幅值。仿真结果表明,所提出的算法明显改进了Burg算法的谱估计性能,能够在噪声环境下获得准确的间谐波参数估计。

The Burg algorithm is suitable for inter-harmonic estimation, but, in the presence of noise, the estimated spectral results will be biased or even erroneous. A new Burg adaptive algorithm based on the higher-order cumulant is put forward to estimate the inter-harmonic signals. First, this paper introduces the principle of Burg algorithm, and modifies the error prediction power of the algorithm based on the fact that higher-order cumulant is insensitive to Gaussian noise. Then, the recursive form is presented to calculate the reflectance so as to reduce the calculation workload brought about by the cumulant. Finally, the fourth-order cumulant diagonal slice of the inter-harmonic signal is proposed to estimate the amplitudes of various components. Simulation results show that the method proposed can markedly improve the spectral estimation performance while yielding accurate inter-harmonic frequency and amplitude even in a noisy environment.