低成本IMU的多位置旋转现场标定方法

针对低成本IMU的系统误差难以现场快速标定问题,提出了一种无需任何外部设备辅助的多位置旋转现场标定方法。该方法通过比力加速度与重力建立加速度计的误差模型,基于动态旋转以及标定后的加速度建立导航方程实现陀螺仪误差建模,使用改进的LM算法,实现低成本IMU误差参数的快速标定。实验结果表明:该方法可以有效地标定出加速度计和陀螺仪的安装误差、缩放因子和零偏误差,极大地简化了标定的过程,标定补偿后的IMU原始数据质量大幅提高,在100 s的静态导航试验中,x、y、z的定位精度分别从2541.547m、895.191m、7267.507m提升至80.229m、41.430m、99.832m。

Method for low-cost IMU in-field calibration through multi-position rotation

To realize the in-field fast calibration of low-cost inertial sensors without using expensive turntable,a multi-position rotation field calibration method is proposed,which does not require any external equipment.The accelerometer error model is constructed by comparing the sensed specific force with the Earth's gravity,and the calibration of gyros is carried out by combining the calibrated accelerations with INS navigation algorithm.An improved Levenberg-Marquardt algorithm is proposed to calibrate the IMU sensor errors.The static navigation test is carried out to test the effectiveness of the proposed algorithm,and the results show that the proposed calibration scheme can be easily implemented to estimate the errors of the accelerometers and gyros,including axis misalignments,scale factors,and biases,and the quality of IMU raw data is significantly improved,in which the positioning accuracies of x/y/z components during 100s static navigation test are improved from 2541.547m/895.191 m/7267.507m to 80.229m/41.430m/99.832m.