一种基于概率数据关联的地磁匹配ICCP算法

Iterated closest contour point algorithm for geomagnetic matching based on probability data association

针对地磁测量误差和地磁补偿余差等带来的匹配误差增大和易造成误匹配的问题,提出了一种基于概率数据关联的地磁匹配ICCP算法。首先,基于磁场测量误差的统计特性,在一定置信范围内生成若干伪测量值,将其视为对应位置地磁场的有效测量,只是它们源于磁场真实值的概率不同;然后,利用ICCP算法对各伪测量值构成的磁场序列分别进行匹配,得到干扰环境下载体可能出现的位置;最后,利用载体的运动约束对各匹配结果进行限制,再对有效的匹配结果进行融合,得到载体最终的位置。实验结果表明,在不同磁场干扰水平下,所提算法的定位精度都能控制在2个地磁图网格左右,具有较好的应用前景。

Aiming at the problem that the magnetic measuring error and geomagnetic compensation residual can increase the matching error and may lead to mismatch, an improved ICCP geomagnetic matching algorithm based on probability data association（PDA） is proposed. Firstly, considering the statistical property of the magnetic measuring error, several pseudo magnetic measurements are generated within a confidence range. They are all considered to be effective measurements of the corresponding position, but the probabilities that they derive from the real magnetic value are different. Secondly, the pseudo measurements are used by the Iterated Closest Contour Point（ICCP） matching algorithm to find the possible positions of the carrier. Finally, carrier＇s kinematical constraints are applied to limit the matching results, and the effective matching results are fused to estimate the final position of the carrier. Experiments results show that the matching errors of the proposed algorithm can all be controlled to about 2 of geomagnetic map grids under different interference levels, which has good application prospect.