IMU姿态解算的卡尔曼滤波方法比较

Comparison of Kalman Filtering Algorithms for IMU Attitude Estimation

【目的】利用惯性定位技术在非开挖时直接测定电缆通道的三维坐标。在惯性定位中,姿态算法对整个系统精度的影响最大,对扩展卡尔曼滤波(EKF)和误差状态卡尔曼滤波(ESKF)两种常用的姿态解算法进行评估。【方法】以四元数和陀螺输出误差为状态向量、磁力计和加速度计为观测向量,基于四元数姿态模型,分别用EKF和ESKF进行姿态解算。【结果】试验结果表明:在计算效率方面,ESKF的计算效率要高于EKF,ESKF的计算耗时约为EKF的80%;在计算精度方面,EKF的均方根误差相对于ESKF有35.6%的改善。【结论】由于地下管线惯性定位主要考虑定位精度,因此EKF具有更好的应用价值。

[Purposes]To directly measure the three-dimensional coordinates of cable channel by inertial positioning technology in trenchless condition.In inertial positioning,the attitude algorithm has the greatest impact on the accuracy of the entire system.Two commonly used attitude solution algorithms,extended Kalman filter(EKF)and error state Kalman filter(ESKF),are evaluated.[Methods]Based on the quaternion attitude model,EKF and ESKF are used to solve the attitude with quaternion and gyro output error as state vectors,magnetometer and accelerometer as observation vectors.[Findings]The experimental results show that the calculation efficiency of ESKF is higher than that of EKF,and the calculation time of ESKF is about 80%of that of EKF.In terms of calculation accuracy,the root mean square error of EKF is improved by 35.6%compared with ESKF.[Conclusions]Since the inertial positioning of underground pipelines mainly considers the positioning accuracy,EKF has better application value.