论文部分内容阅读
提出了考虑阵元互耦的子空间拟合波达方向估计方法。用一带状、对称Toeplitz矩阵对均匀线阵的互耦矩阵建模。根据互耦矩阵自由度,去除部分两边阵元,研究了子空间拟合算法中ESPRIT算法的理论性能。在大快拍的情况下,均匀线阵的MCLS-ESPRIT算法的角度估计误差是一个零均值的联合高斯分布,同时给出了分布的协方差表达式。由方差的数值计算可知,在阵元数和信源数确定的情况下,存在估计标准差最小的子阵;与不存在互耦的理想情况相比,除了估计标准差少许增加外,性能接近。本文的方法已应用于高分辨率测深侧扫声呐,获得了好结果。在声呐阵阵元数增加的条件下,用本文的方法,能抑制均匀线阵的互耦的影响。
A subspace fitting DOA estimation method considering the mutual coupling of array elements is proposed. Modeling a Mutual Coupling Matrix of Uniform Linear Arrays with a Banded and Symmetric Toeplitz Matrix. According to the degree of freedom of the mutual coupling matrix, some of the elements on both sides are removed, and the theoretical performance of the ESPRIT algorithm in the subspace fitting algorithm is studied. In the case of large snapshots, the angular estimation error of the uniform linear array MCLS-ESPRIT algorithm is a zero-mean joint Gaussian distribution, while the distribution of the covariance expression is given. From the numerical calculation of variance, we can see that in the case of determining the number of elements and the number of sources, there is a sub-array with the smallest estimated standard deviation. Compared with the ideal without mutual coupling, the performance is close to the standard one . The method in this paper has been applied to high-resolution sounding side-scan sonar, and good results have been obtained. With the increase of sonar array element number, the influence of the mutual coupling of uniform linear arrays can be suppressed by this method.