基于二阶插值滤波的室内航向估计算法研究

发布时间：2018-03-14 03:23

本文选题：室内导航　切入点：插值滤波　出处：《压电与声光》2017年03期 　论文类型：期刊论文

【摘要】：在室内行人导航中,为获取精确的航向角,提出了一种基于二阶插值滤波的航向估计算法。该算法通过建立四元数姿态运动测量模型,采用二阶插值滤波对陀螺仪、加速计和磁强计的测量数据进行数据融合,实现航向角解算,以过程噪声和测量噪声为设计参数,构造自适应噪声协方差矩阵,实现协方差误差估计最小化。通过对行走长方形参考路径得到的数据进行处理,加速计和磁强计组合与陀螺仪航向估计算法动态误差分别为13.6°和6.9°,采用二阶插值滤波航向估计算法后动态误差为2.3°。实验结果表明,该算法有效提高了航向估计的精度,减小了陀螺仪漂移、载体的线性加速度和周围局部磁场干扰对航向估计的影响。
[Abstract]:In order to obtain accurate heading angle in indoor pedestrian navigation, a heading estimation algorithm based on second-order interpolation filter is proposed, which establishes quaternion attitude motion measurement model and adopts second-order interpolation filter for gyroscope. The measurement data of accelerometer and magnetometer are fused to realize the calculation of heading angle. Taking process noise and measurement noise as design parameters, the adaptive noise covariance matrix is constructed. The covariance error estimation is minimized by processing the data obtained from the walking rectangular reference path. The dynamic errors of the accelerometer and magnetometer combined with gyroscope are 13.6 掳and 6.9 掳, respectively, and the dynamic error is 2.3 掳after using the second-order interpolation filtering algorithm. The experimental results show that the algorithm can effectively improve the accuracy of the course estimation. The gyroscope drift, the linear acceleration of the carrier and the interference of the surrounding local magnetic field on the course estimation are reduced.
【作者单位】：西安邮电大学电子工程学院;
【基金】：陕西省科技统筹创新工程计划资助项目(2014KTCQ01-21) 西安邮电大学研究生创新基金资助项目(CXL2015-38)
【分类号】：TN96

【参考文献】