利用信号频谱幅度分布的先验信息 ,对短时数据序列信号提出了一种实现高分辨率谱估计和数据序列外推的快速算法。该算法利用了算子矩阵的特殊结构 ,采用简单、快速的矩阵求逆算法 ,使算法的运算量和数据存储容量大幅度下降。给出了算法在高分辨率谱估计及数据序列外推中的应用实例和仿真结果 ,证实了该算法的优良性能。 Using the prior information of the amplitude distribution of the signal, a fast algorithm for high resolution spectral estimation and data sequence extrapolation is proposed for the short-time data sequence signal. The algorithm takes advantage of the special structure of the operator matrix and adopts a simple and fast matrix inversion algorithm, which greatly reduces the computational complexity and data storage capacity of the algorithm. The application examples and simulation results of the algorithm in high resolution spectral estimation and extrapolation of data sequence are given, and the good performance of the algorithm is verified.