中文摘要：

核四极矩共振（NQR）是一种固态射频谱分析技术，可用于检测高危险爆炸物．然而NQR信号本身非常弱，并且易受线圈的热噪声和外部射频干扰的影响，低信噪比限制了NQR的实际应用．该文提出一种改进的微弱NQR信号检测算法．首先利用Hankel矩阵方式下奇异值分解的方法，有效地抑制射频干扰和噪声，并将NQR信号分离出来．然后提出了一种基于MUSIC谱估计的非线性最小二乘检测器，它既保证了高的频率分辨率，又大大降低了运算量．仿真数据和实测数据结果表明该算法的有效性．

英文摘要：

Nuclear quadrupole resonance （NQR） is a solid-state radio frequency （RF） spectroscopic technique allowing the detection of many high explosives. However, the NQR signals are intrinsically weak, and the practical use of NQR is hindered by low sig-nal-to-noise ratio （SNR）, and artifacts arising from the thermal noise of the coil and ex-ternal radio frequency interference （RFI）. In this paper, we described an improved de-tection algorithm for detection of weak NQR signals. First, singular value decomposi-tion using the Hankel matrix was utilized to restrain RFI and filter noise. A non-linear least squares detector based on the （multiple signal classification） MUSIC spectral esti- mation was applied, which not only guaranteed a high resolution in the frequency do-main, but also led to significant reduction in the amount of computation. The effective-ness of this algorithm was demonstrated with the simulated and experimental results.