快速ESPRIT谐波集检测算法

Improvement of ESPRIT algorithm for the harmonic set detections

在各种高分辨谐波检测算法中,TLS-ESPRIT估计精度较高,因不需要谱峰搜索而具有较小的运算量。提出了一种快速ESPRIT算法,在不降低谱估计性能条件下,进一步减小了运算量。快速ESPRIT算法所需采样时间短,仅需两次特征值分解。分析表明,此算法可以检测不同飞行状态的直升机声信号谐波集,是一种实时性好、有效实用的目标谱集探测方法。

In a variety of high-resolution harmonic detection algorithms, TLS-ESPRIT has higher the estimation precision, and less computation because it needn＇t peak searching. A fast ESPRIT algorithm is proposed in this paper, and it can further reduce the computational complexity without reducing the spectral estimation performance conditions. The fast ESPRIT algorithm required sampling time is short, and it only needs two eigenvalue decomposition. Our tests on the real data show that the method has good performance to the helicopter in almost all cases, and it can extract effective features for the helicopter target identification furthermore.