【摘要】 针对现有波达方向估计(Direction of Arrival, DOA)算法在低信噪比、多信源条件下估计精度不足、效率低等问题，提出了一种基于可分离替代函数算法的矢量水听器阵列多快拍DOA估计方法.首先对空域等角度均匀划分，构造出超完备冗余字典，建立信号多快拍数据在空域的稀疏表示模型，然后采用可分离替代函数算法思想解决稀疏重构问题，求解出信号在空域的稀疏系数矩阵，最后将稀疏矩阵中行向量的范数映射到空域网格上，得到DOA估计值.仿真实验表明：该方法在低信噪比、多信源条件下拥有比子空间类算法、贪婪类算法以及现有凸优化类估计算法更高的DOA估计精度和更强的鲁棒性，与同类算法相比执行效率更高.更多还原

【Abstract】 Aiming at the problems of the existing direction of arrival(DOA) estimation algorithms, such as insufficient estimation accuracy and low efficiency under the conditions of low SNR and multiple sources, a multi snapshot DOA estimation method for vector hydrophone array based on separable surrogate function algorithm is proposed.First, the spatial domain is evenly divided from equal angles, a super complete redundant dictionary is constructed, and the sparse representation model of multi snapshot data of signals in the spatial domain is established.Then, the separable replacement function algorithm is used to solve the sparse reconstruction problem, that is, the distance metric function with the previous solution is added to the original problem, and the contraction operation is used to accelerate the convergence in the iterative solution process, so as to solve the sparse coefficient matrix of signals in the spatial domain, Finally, the norm of the row vector in the sparse matrix is mapped to the spatial grid to obtain the DOA estimate.The simulation results show that this method has higher DOA estimation accuracy and stronger robustness than subspace algorithm, greedy algorithm and existing convex optimization algorithm under the condition of low SNR and multiple sources, and has higher execution efficiency compared with similar algorithms.更多还原