基于行人航位推算的后向地磁匹配算法

针对行人航位推算(pedestrian dead reckoning,PDR)定位存在误差累积和智能手机内置传感器精度不高的问题,采用PDR结合地磁的方法进行室内定位研究,提出一种改进的基于PDR的后向地磁匹配算法.在构建地磁基准库阶段,使用克里金插值算法有效减少数据采集所耗费的大量时间并构建出双分辨率地磁基准库.在地磁匹配阶段,改进了基于动态时间规整(dynamic time warping,DTW)的后向地磁匹配算法,改进后的算法避免传统DTW地磁匹配需要全局搜索地磁序列的缺点,在保证定位精度的前提下,增强了定位实时性.实验结果表明,本文算法最大定位误差小于1.5m,可以满足普通室内定位需求.

Backward Geomagnetic Matching Algorithm Research Based on Pedestrian Dead Reckoning

Aiming at the problems of error accumulation in pedestrian dead reckoning (PDR) positioning and low precision of sensors embedded in smart phones, this paper proposes to study indoor positioning using PDR combined with geomagnetic method. In the construction of the geomagnetic reference library, the kriging interpolation algorithm is used to reduce the large amount of time spent in data acquisition and to construct the dual-resolution geomagnetic reference library. In the phase of geomagnetic matching, this paper proposes a backward region constraint matching algorithm based on dynamic time warping (DTW). This algorithm avoids the shortcoming of global search of geomagnetic sequence required by traditional DTW geomagnetic matching, and enhances the real-time positioning on the premise of ensuring the positioning accuracy.