摘要
针对可分离分布平面阵列的稀布优化问题,提出了一种基于矩阵束方法的减少阵元数目、求解阵元位置和设计激励幅度的优化方法。可分离分布平面阵的方向图等于两个正交线阵方向图的乘积。对形成期望方向图的两正交线阵的方向图进行采样得到离散的数据集,再构造Hankel矩阵;然后对此Hankel矩阵进行奇异值(SVD)分解,舍弃一部分不重要的奇异值,得到近似Hankel矩阵的最优的低秩逼近矩阵,它和稀布线阵的方向图相一致;基于广义特征值分解的最小二乘准则来计算两稀布线阵的阵元位置和激励,从而得到稀布面阵的位置和激励。仿真结果证实了该算法的有效性。
For the optimization of sparse separable array, a new method based on matrix pencil method is proposed to reduce the number of elements, to solve the element locations and to design the excitations. The pattern of seperable planar array is produced by the pattern of two orthogonal linear array. Firstly,the two desired orthogonal linear array radiation pattern is sampled to form a discrete data set. Secondly,a Hankel Matrix is built and the singular value decomposition （SVD） can be performed. By discarding the insignificant singular values, we obtain an optimal lowrank approximation of the Hankel matrix which corresponds to sparse antenna array. Finally,the generalized eigen-decomposition is employed to calculate the sparse linear array locations and excitations, then the locations and excitations of the sparse separable array are obtained. Simulation results are presented to demonstrate the efficiency of the proposed approach.
出处
《电波科学学报》
EI
CSCD
北大核心
2012年第6期1180-1185,1231,1267,共8页
Chinese Journal of Radio Science
基金
国家自然科学基金资助项目(No.60736045)
中国博士后科学基金(2012M511919)
