三种基于四元数模型的声矢量阵MUSIC算法  

Three MUSIC Algorithms Based on Quaternion Model of Vector Hydrophone Array

作　　者：何光进 高峰 He Guangjin;Gao Feng(The Third Military Representative Office in Guangzhou, Guangzhou 510000, China)

机构地区：[1]驻广州地区第三军事代表室

出　　处：《船电技术》2019年第10期28-32,共5页Marine Electric & Electronic Engineering

摘　　要：传统的基于长矢量模型的MUSIC算法没有充分利用矢量水听器各分量之间的正交特性,存在多目标分辨能力不足的缺点。针对这一情况,文中引入了四元数的概念,利用四元数能更好地描述矢量水听器各分量的正交结构这一优点,提出了三种基于四元数模型的二维声矢量阵方位估计算法:Q-VV算法、Q-PV算法和Q-PVV算法。与基于长矢量模型的MUSIC算法相比,新算法降低了协方差矩阵的维数,占用更少的内存空间。同时,由于利用了四元数强正交性约束的特点,新算法有好的单目标估计精度和多目标分辨能力。其中:Q-PV算法和Q-PVV算法利用了声压与振速的相干性,抗各向同性干扰的能力优于Q-VV算法。仿真实验证明了算法的有效性。In the vector hydrophone processing, the traditional MUSIC algorithm based on long-vector approach can’t take full use of the orthogonality of the velocity information. The ability to distinguish multiple targets is weak. In response to this situation, the quaternion algebra is introduced since the orthogonality of quaternion is suitable for describing the orthogonal structure of vector hydrophone. Three Direction-Of-Arrival (DOA) algorithms (Q-VV, Q-PV and Q-PVV) of 2-D vector hydrophone array are proposed. Compared with the long-vector MUSIC method, each of the algorithms takes up less memory space due to the reduction of covariance matrix dimension. As the stronger orthogonality constraints, these algorithms have a higher ability of multiple targets resolution. Wherein, algorithm Q-PV and Q-PVV are robust to isotropic interference by taking use of the coherence of pressure and velocity. The simulation proves the merits mentioned above.

关 键 词：声矢量阵 MUSIC算法 四元数 方位估计 联合方位估计 

分 类 号：TB566.21[交通运输工程—水声工程]