特征矢量法是分辨相关信号源的一种有效算法。该算法需求出阵列接收信号空间相关矩阵的最小特征矢量。本文在梯度法与空间三角格型法基础上,提出了一种新的算法:变换自适应法。此法通过一个变换,将求矩阵最小特征矢量的问题转换成求最大特征矢量,从而避免了矩阵求逆这一繁杂过程。理论分析以及计算机仿真结果均表明,该算法在计算量和收敛速度方面具有相当的优越性。 Eigenvector method is an effective algorithm to distinguish the relevant signal source. The algorithm requires the minimum eigenvector of the spatial correlation matrix of the received signal of the array. In this paper, based on the gradient method and the triangular space method, a new algorithm is proposed: transform adaptive method. This method transforms the problem of finding the smallest eigenvector of the matrix into the largest eigenvector through a transformation, thus avoiding the complicated process of matrix inversion. Theoretical analysis and computer simulation results show that the proposed algorithm has considerable advantages in terms of computational complexity and convergence speed.