基于SVD-TLS的AR谱估计方法在声目标识别中的应用

为提高 AR谱估计方法的抗噪性能 ,提出采用 SVD-TLS算法。文中对该算法作了介绍 ,并用一简单例子将 SVD-TLS与其它算法作了比较 ,最后应用 SVD-TLS对几组实测的直升机声信号进行了 AR参数估计和谐波频率提取。仿真结果表明 SVD-TL S较之其它算法可在更低的信噪比情况下准确地提取信号的频率信息 ,具有良好的抗噪性能和数值稳定性 ,将其用于声目标识别对改善系统的抗噪性能是十分有意义的。

The Application of AR Spectrum Estimating Method Based on SVD-TLS in Acoustic Target Identification

To improve the anti noise performance of AR spectrum estimating method, the algorithm based on SVD TLS is introduced and compared with other ones with the help of an example in this paper. Finally, AR parameter estimation of the acoustic signals measured from a helicopter is made and harmonic frequencies are extracted by using SVD TLS. The simulation results have shown that compared with other algorithms, SVD TLS can extract the frequency information of the signals precisely under the condition of lower signal to noise ratio. This method is possessed of good anti noise performance and numberstability, and can be used to do acoustic target identification to improve the anti noise performance of the system.