空间谱估计技术具有波达方向估计精度高、分辨能力强、并且可以同时估计多个信号等特点。但是阵列误差会导致实际的阵列流型和理论值存在偏差，这会使得空间谱估计技术性能下降甚至失效，因此阵元误差是空间谱估计技术走向实用的瓶颈。互耦误差相对其它误差形式是一种未知参数较多，校正方法比较复杂的误差形式，阵列互耦误差校正一直很难找到一个简单且有效的方法，因此互耦误差校正方法也就成为当前研究的热点问题之一。 本文针对上述问题，较深入的分析和研究阵列互耦误差的校正方法。文中从空间谱估计角度介绍了阵列误差校正方法的基础，介绍了基于MUSIC算法的阵列互耦误差的模型及互耦误差对算法的影响。并将阵列误差校正方法分为两个大类：一类通过估计互耦误差的方法实现误差校正，包括有源校正方法和自校正方法。该方法主要是对阵列互耦误差进行数学建模，然后将互耦误差校正问题转化为多参数联合估计问题，进而寻求最优解实现误差估计，最后通过补偿这些阵列误差完成互耦校正。另一类不需要估计阵列误差，采用算法自身对阵列误差的不敏感特性来实现良好的误差估计，即稳健算法。分析和研究了 和 型阵列的数学模型及结构特性，利用其对称结构和互耦特性给出以自校正算法为基础的 阵列互耦校正算法，该算法避免了多维搜索带来的庞大运算量和迭代中的全局收敛性问题，不需要校正源就可以实现两类参数的估计。相对迭代算法，该算法具有运算量小、估计精度高的特点。最后通过仿真实验，验证了各个校正方法的互耦误差估计性能。

Spatial spectrum estimation technique has a high DOA estimation accuracy, the ability to distinguish, and can estimate the characteristics of multiple signals. Array error will result in the actual array flow pattern and the theoretical value of the deviation, which would make the space spectral estimation technology performance degradation or even failure, so the element error is the spatial spectrum estimation technique toward practical bottleneck. Relative to other errors in the form of mutual coupling error is an unknown parameter and the more complex error correction method, the array mutual coupling error correction has been difficult to find a simple and effective method of mutual coupling error correction method, so mutual coupling error has become one of the current studying hotspots. This paper focus on the above problem deeply analyzed and studied the correction method of array mutual coupling error. This article introduced the foundation of the array error calibration method from the perspective of spatial spectrum estimation, and introduced the MUSIC algorithm based on array mutual coupling error model and error on the effects of mutual coupling algorithm. Array calibration method is divided into two categories: a category error by estimating the mutual coupling method to achieve error correction, including active correction method and self-correction method. The method is mainly to the array mutual coupling errors in mathematical modeling, then mutual coupling error correction into a multi-parameter joint estimation problem, and then finding the optimal solution to achieve the error estimates, and finally completed by these arrays error compensation of mutual coupling correction. The other do not need to estimate the error of the array, using the algorithm itself on the insensitivity of the array error to achieve good error estimates, that is robust algorithm. Analysis and research on the L-array and Y-array of mathematical models and structural, and according to the properties of symmetry and mutual coupling characteristics, the paper are given to self-correction algorithm based on the L-array mutual coupling correction algorithm, the algorithm avoids the multidimensional search for the enormous computational and global convergence of the iteration, and does not require calibration source can be of two types of parameters estimated. Relative iteration algorithm, it is characteristic of high accuracy and little computation. Finally, through simulation experiments to verify the mutual coupling error correction method estimation performance.