基于PDR和磁场匹配的智能手机室内定位

目前,基于磁场的室内定位方法存在指纹采集无法快速建库的问题,导致匹配特征较少、无法快速进行匹配操作和航向角估计精度低。基于磁场梯度的快速近似,使用高斯牛顿迭代方法结合通过PDR测量的轨迹进行磁场轮廓匹配定位,提高了单点磁场指纹的可分辨性。基于步态模型的更新用作测量信息以改善C-INS的导航性能,提出了一种基于捷联惯性导航系统的行人航位推算算法。基于MFS参考位置,使用扩展Kalman滤波器控制PDR惯性导航的位置漂移误差,进一步提高了轨迹轮廓的精度。测试结果表明,该算法可以获得更好的位置估计和航向估计结果。

Indoor Location of Smartphone Based on PDR and Magnetic Field Matching

At present, the method of indoor location based on magnetic field has the problem that fingerprint acquisition method can''t build database quickly, which leads to less matching features, slow matching operation and low accuracy of angle estimation. Based on the fast approximation of magnetic field gradient, the Gauss Newton iterative method combined with the trajectory from PDR measurement is used to match and locate the magnetic field contour, so as to improve the resolution of single point magnetic field fingerprint. Based on the gait model updating as measure-ment information to improve the navigation performance of C-INS, a pedestrian dead reckoning algorithm based on strapdown inertial navigation system (SINS) is proposed. The MFS reference position is used to control the position drift error of PDR inertial navigation, and the accuracy of trajectory contour is further improved. The test results show that the algorithm can obtain better position estimation and heading estimation results.