基于地磁辅助的室内行人定位航向校正方法

发布时间：2018-12-16 12:51

【摘要】：在基于惯性传感器的室内人员自主定位系统中,由于陀螺仪存在随时间增长的漂移,需要依靠地磁场对航向校正。然而,室内场景中地磁场会受到比较严重的干扰,磁力计自身也存在一定误差,严重影响了地磁场辅助定位的航向校正效果。该文提出一种地磁辅助的室内定位航向校正方法,该方法首先结合人体运动模型对磁力计进行校准,获取相关校准系数,提高磁力计解算航向精度。在此基础上利用改进的磁场准静态检测方法提取可用的磁力计观测量,经校准系数校准后结合零速更新算法(Zero velocity UPda Te,ZUPT)辅助的扩展卡尔曼滤波(Extended Kalman Filter,EKF),对陀螺仪航向进行滤波校正。实验结果显示在长约684 m的轨迹下航向误差仅为3.2°定位误差小于0.5 m,定位精度优于0.2%,验证了该方法较传统的磁场辅助航向校正方法的优越性。
[Abstract]:In the indoor autonomous positioning system based on the inertial sensor, the gyroscope needs to rely on the geomagnetic field to correct the course because of the drift of the gyroscope with the increase of time. However, the geomagnetic field will be seriously disturbed in indoor scene, and there are some errors in the magnetometer itself, which seriously affects the course correction effect of geomagnetic field assisted positioning. In this paper, a geomagnetic assistant method for heading correction in indoor positioning is proposed. Firstly, the magnetometer is calibrated with human motion model, and the correlation calibration coefficient is obtained to improve the precision of course calculation. Based on the improved quasi-static magnetic field detection method, the available magnetometer observations are extracted, and the extended Kalman filter (Extended Kalman Filter,EKF (Extended Kalman Filter,EKF), which is aided by the zero-velocity update algorithm (Zero velocity UPda Te,ZUPT), is calibrated after calibrating the magnetometer parameters. The gyroscope course is filtered and corrected. The experimental results show that the heading error is only 3.2 掳and the positioning accuracy is less than 0.2 m under the trajectory of about 684m, which verifies the superiority of this method compared with the traditional method of magnetic field aided course correction.
【作者单位】：国防科技大学电子科学与工程学院;
【分类号】：TN96

【相似文献】