被动声传感器网络中的一种去除虚假定位点的新方法

被动声传感器网络中多目标跟踪经常会出现错误关联(虚假定位点)。为了去除该定位点，提出了一种基于广义三角定位去除问题的新方法。利用灰色关联理论对声探测阵接收到的不同目标的乱序角度信息进行重新排序，获得正确的角度次序。通过基于角度关联预处理算法的时段匹配三角定位法获取不同目标最可能的位置。然后利用本文提出的假设和第三站验证方法对虚假定位点进行剔除。仿真数据和实际试验数据验证了所提出方法的有效性。

A new algorithm for deghosting in passive acoustic sensor network

This paper addresses the false association (called ghosts) problem of multi-target tracking in a distributed passive acoustic sensor network. To eliminate these ghosts, a new deghosting scheme based on the generalized triangulation has been proposed. First, the received angle information of different targets is reordered by using the gray theory to obtain the correct sequence. Successively, time span match triangulation based on pretreatment algorithm of angle association is used to get the most-likely positions of different targets. Finally, the residual ghosts are cancelled by using a reasonable hypothesis and third site notarization. The validity of the proposed scheme is evaluated by both simulation and experimental results.