该文将高阶统计量应用于线性、时不变非最小相位ＦＩＲ系统的识别问题。对由非高斯、非对称分布白噪声激励的非最小相位ＦＩＲ系统，当仅已知其输出信号（可能含有噪声）时，该文提出一种利用输出信号三阶、四阶累积量进行系统辨识的新算法。在计算机仿真中，将该算法与ＧＭ法和Ｔ法进行比较，结果表明，当输出信号中含有高斯有色观测噪声时，新算法的处理效果明显好于ＧＭ法和Ｔ法 In this paper, high order statistics are applied to the identification of linear, time invariant non-minimum phase FIR systems. For non-minimum phase FIR systems excited by non-Gaussian and asymmetric white noise, when only the output signal (which may contain noise) is known, this paper presents a system identification using the third and fourth order cumulants of the output signal New algorithm. In computer simulation, the algorithm is compared with the GM method and the T method. The results show that when the output signal contains Gaussian colored noise, the new algorithm is better than GM method and T method