论文部分内容阅读
为了解决现有阵列失效单元诊断方法中存在的采样数量大、测量时间长、诊断效率低等缺陷，提出了一种采用非均匀随机欠采样策略的压缩感知诊断新方法。该方法基于失效单元数目的稀疏性，通过完好阵列和实际阵列激励的差值构造稀疏信号，并根据目标出现在空间的不同方位选择相对最优的采样策略，进而构造对应的测量矩阵获取空间测量数据。在此基础上，利用平行坐标下降算法精确重构该稀疏信号，实现了对失效单元的准确诊断。理论分析和仿真实验表明，本文提出的方法不仅明显减少了采样数量，有效缩短了诊断时间，显著降低了计算复杂度，而且进一步提高了所重构辐射特性的精确度。
In order to solve the defects of large number of samples, long measuring time and low diagnostic efficiency, the new method of compressed sensing based on non-uniform random under-sampling strategy is proposed. Based on the sparsity of the number of failed cells, a sparse signal is constructed based on the difference between the perfect array and the actual array excitation, and a relatively optimal sampling strategy is selected according to the different azimuths of the target appearing in the space. The corresponding measurement matrix is constructed to obtain the spatial measurement data. On this basis, the sparse signal is accurately reconstructed by the parallel coordinate descent algorithm, and the accurate diagnosis of the failed unit is realized. Theoretical analysis and simulation experiments show that the proposed method not only reduces the number of samples significantly, shortens the diagnosis time, significantly reduces the computational complexity, but also further improves the accuracy of the reconstructed radiation characteristics.