基于压缩感知的区域离散化矿井目标定位方法

Region discretization mine target positioning method based on compressed sensing

  • 摘要: 针对电磁波测距定位方法受井下电磁干扰导致定位精度低,指纹定位方法离线构建指纹特征库面临海量离线数据连续测量和存储的问题,提出了基于压缩感知的区域离散化矿井目标定位方法。首先对定位区域划分网格进行离散化处理,离线测量网格处信号强度,构建指纹特征库;然后将在线采样的少量信号强度通过压缩感知得到包含目标位置信息的稀疏位置矩阵,从而估计目标可能存在的位置;最后通过网格质心法对目标精确定位。实验结果表明,该方法提高了定位精度、缩短了定位时间,实现了矿井目标精确定位。

     

    Abstract: For problems of low positioning accuracy of electromagnetic wave ranging positioning method due to underground electromagnetic interference and continuous measurement and storage of massive offline data in offline construction of fingerprint feature database by fingerprint positioning method, a region discretization mine target positioning method based on compressed sensing was proposed. Firstly, positioning area is divided into grids for discretization, and fingerprint feature database is constructed by offline measuring signal strength at the grid. Then, sparse position matrix containing target position information is obtained by compressed sensing with a small amount of signal strength sampled online, so as to estimate possible position of the target. Finally, the target is accurately positioned by grid centroid method. The experimental results show that the method achieves accurate positioning of mine target with higher positioning accuracy and less positioning time.

     

/

返回文章
返回