基于贝叶斯估计的浅海声源被动定位方法

Passive Localization Method for Acoustic Sources in Shallow Water Based on Bayesian Estimation

针对浅海环境干涉复杂和非稳定特性,建立概率密度函数形式的声场模型,有效克服定位时的模型失配问题;设计以声源状态矢量为后验概率的贝叶斯定位模型,用迭代形式达到时间换空间的目的,实现单个水听器对运动目标的定位;提出网格划分直方图滤波法,将解析积分转化成数值求和,提高算法效率。SWellEx-96实测数据验证结果表明,深度200 m、距离10 km的范围内,深度定位相对误差可控制在12.04%,距离定位相对误差可控制在6.47%。新方法可用于浅海隐蔽低耗武器平台对移动目标的探测。

An acoustic field model in the form of probability density function is established to solve the problem of model mismatch in shallow water.The Bayesian localization model with the state vector of an acoustic source as a posteriori probability is designed to achieve the purpose of exchanging time for space by iteration and realize the localization of a moving target by a single hydrophone.The grid histogram filtering algorithm is proposed to convert analytical integral into numerical summation and improve efficiency of the algorithm.The SWellex-96 experimental results show that the relative error of depth localization can be controlled at 12.04%and the relative error of distance localization at 6.47%within a depth of 200 m and a distance of 10 km.The proosed method can be used to detect targets in shallow water with concealed low-energy consumption weapon platform.