被动声纳方位估计的宽带盲源分离算法

针对实际被动声纳信号宽带非平稳且统计特性无法预知的特点，由宽带卷积混合模型，建立了融合时间延迟结构与非参数化特性的代价函数，通过核密度技术同时估计目标源的概率密度函数和解混矩阵，并对估计的最优解混矩阵与目标源信号求取每个频点内方位能量谱，最后累加所有子带构成宽带方位能量谱。宽带仿真结果与实际海试表明本文方法在方位分辨率和估计精度方面接近最小方差无失真响应（Minimum variance distortionless response, MVDR）和多重信号分类（Multiple signal classification, MUSIC）算法，在弱目标检测方面具有一定优势。

Broad-band blind source separation algorithm for passive sonar bearing estimation

Aiming at the particular characteristics of real passive sonar signal that is of broad-band,nonstationary and little knowledge of statistics,a fusion cost function was established which combines the property of time delay structure with that of non-parametrization in the light of broad-band convolution mixing model.Using a non-parametric kernel density estimation technique,the algorithm performs simultaneously the estimation of the probability density functions of the source signals and the estimation of the unmixing matrix.And then,the proposed algorithm uses the optimized unmixing matrix and estimated target source signals to achieve a bearing-energy spectrum component in each frequency bin which is accumulated to obtain the broad-band bearing-energy spectrum.Broad-band simulation results and real sea trial show that the results of the proposed method are close to those of the minimum variance distortionless response(MVDR) and multiple signal classification(MUSIC) methods in the aspect of bearing resolution and estimation accuracy,and moreover,play better in the aspect of weak target detection where strong interference exists.