基于重复广义互相关时延估计的超短基线水声定位算法

Ultra-short baseline underwater acoustic location based on repeated generalized cross-correlation time delay estimation

针对在超短基线水声定位中,传统时延估计算法抗噪声性能差的问题,提出了一种基于重复广义互相关时延估计的超短基线水声定位算法。通过对两个源信号的其中一个增加一次自相关后,再对这两个源信号进行合适的加权广义互相关,之后对反傅里叶变换后得到的广义互相序列进行峰值检测,最终得到时延。通过半物理实验,比较经典的互相关算法和重复广义互相关算法发现,在相同环境中,新提出的算法时延估计误差为0.0112 s,相对于经典算法的0.0209 s,能够提高46.4%。基于重复广义互相关的超短基线定位算法仿真实验,表明X轴定位精度提高97.1%,Y轴定位精度提高96.3%。

Aiming at the poor anti-noise performance of traditional time-delay estimation algorithm in ultra-short baseline underwater acoustic positioning, an ultra-short baseline underwater acoustic positioning algorithm based on repetitive generalized cross-correlation time-delay estimation is proposed. The algorithm adds one autocorrelation to one of the two source signals, then carries out appropriate weighted generalized cross-correlation to the two source signals, and then detects the peak value of the generalized cross-correlation sequence obtained by inverse Fourier transform, and obtains the time delay. Semi-physical experiments show that the error of the new algorithm is 0.0112 s, and compared with classical cross-correlation algorithm and repetitive generalized cross-correlation algorithm in the same environment, the precision is increased by46.4%, since the error of the classical algorithm is 0.0209 s. The simulation experiments of ultra-short base line location algorithm based on repetitive generalized cross correlation show that the X-axis accuracy and Y-axis accuracy of the new algorithm are improved by 97.1% and 96.3%, respectively.