本文提出一种利用阵列接收信号数据矩阵奇异值分解的时一空二维信号谱估计方法，该方法对线阵接收信号构成的数据矩阵进行奇异值分解，构造出噪声奇异向量，利用信号空间与噪声空间的正交性获得时一空二维谱。该方法可用于高分辨测向测速系统中。计算机模拟表明了该算法的有效性。 In this paper, we propose a time-space two-dimensional signal spectrum estimation method based on singular value decomposition of array received signal data matrix. The method decomposes the data matrix formed by the received signal of the array into singular values ​​and constructs a noise singular vector. The signal space and noise The orthogonality of space gets the empty two-dimensional spectrum. The method can be used in high resolution directional velocity measurement system. Computer simulation shows the effectiveness of the algorithm.