基于改进粒子群算法的矢量水听器多目标方位估计

Multi-target Azimuth Estimation of Vector Hydrophone Based on Improved Particle Swarm Optimization Algorithm

摘要:

为了研究单矢量水听器多目标方位估计能力，本文分别利用互谱声强法、MUSIC算法以及信号统计量方法对多个目标方位进行估计。互谱声强法可以估计出多个不同频的单频目标方位，但对于频谱混叠的目标无法分辨；MUSIC算法可以分辨单频和宽带目标，但利用单矢量水听器最多可估计两个目标方位。为此，本文针对于提出的信号统计量方法，构建了声压和振速的统计量模型，将其与粒子群优化算法及改进算法相结合，实现了基于改进粒子群算法的多目标方位估计。对多个单频和宽带信号目标进行仿真分析，结果表明改进粒子群算法具有良好的估计效果，并对三种方法的估计结果进行比较，验证了改进粒子群算法有较好的适用性。通过对2022年千岛湖试验数据的处理再一次验证了算法的有效性。

Abstract:

In order to study the multi-target azimuth estimation capability of single vector hydrophone, this paper uses cross-spectrum sound intensity method, MUSIC algorithm and signal statistics method to estimate multiple-target azimuth. The cross-spectrum sound intensity method can estimate the azimuth of multiple single-frequency targets with different frequencies, but can not distinguish the targets with spectral aliasing. The MUSIC algorithm can distinguish both single-frequency and wide-band targets, but it can estimate up to two target azimuths with a single-vector hydrophone. Therefore, in this paper, based on the proposed signal statistics method, the statistical models of sound pressure and vibration velocity are constructed, which are combined with particle swarm optimization algorithm and improved algorithm to realize the multi-target azimuth estimation based on improved particle swarm optimization algorithm. The simulation analysis of several single frequency and broadband signal targets shows that improved particle swarm optimization algorithm has a good estimation effect, and the estimation results of the three methods are compared to verify that improved particle swarm optimization algorithm has a good applicability. The validity of the algorithm is verified once again by processing the data of the 2022 Qiandao Lake experiment.