用于雷达方位超分辨的约束迭代Tikhonov正则化算法

Constrained Iterative Tikhonov Regularization Algorithm for Radar Azimuth Super Resolution

摘要:

针对Tikhonov正则化算法噪声适应能力差和不便于引进额外的约束对解进行限制等缺点,利用迭代Tikhonov正则化算法对雷达方位超分辨进行研究,证明了迭代Tikhonov正则化反卷积公式的收敛性,分析了迭代Tikhonov正则化算法的频域性质,在存在噪声的情况下,提出相应的噪声抑制方法,得到约束迭代Tikhonov正则化算法。针对不同信噪比情况进行了计算机仿真实验,结果表明,与迭代Tikhonov正则化算法相比,约束迭代Tikhonov正则化算法具有较强的噪声适应能力,与常用的约束迭代方法(CID)算法相比,具有较快的收敛速度,初步验证了算法的有效性。

Abstract:

An iterative Tikhonov regularization algorithm was used based on Tikhonov regularization algorithm, because of its poor noise adoptive ability and without introduction of additional constraints on the solutions. The convergence of the deconvolution formula was proved, the frequency domain characteristics was analyzed, the corresponding noise suppression methods was given, and the constrained iterative Tikhonov regularization algorithm was got. According to the different signal-to-noise ratio (SNR), the computer simulation experiment was made. The result showed that, compared with iterative Tikhonov regularization algorithm, it had a strong noise adoptive ability, and it had a faster convergence speed, which verified the algorithm efficiency preliminarily.