MEMS惯性传感器误差简易标定方法

A simple calibration method of MEMS inertial sensor error

针对传统MEMS惯性传感器标定方法依赖于专业外界设备及复杂的标定流程的问题,提出一种基于IMU原地旋转的加速度计、陀螺仪零偏和标度因数同步估计的简易标定方法。简化惯性解算微分方程与误差模型,利用传感器误差与导航参数误差耦合的特性,仅依靠重力加速度为参考信息设计IMU旋转方法;并根据IMU原地旋转条件下速度、位置变化几乎为零的特征,结合惯性解算速度、位置信息设计伪观测量,构建卡尔曼滤波估计传感器误差;最后分别开展了转台模式和手持模式标定实验。实验结果表明,只需手持IMU作有效旋转即可在1min内得到收敛的MEMS加速度计、陀螺仪零偏和标度因数。

Aiming at the problem that the traditional calibration methods of MEMS inertial sensors rely on professional equipment and complex procedures, the paper proposed a simple synchronous calibration method of bias and scale factors of accelerometers and gyros based on the in situ rotations of inertial measurement unit（IMU）. The IMU rotation method was designed with the only referencial information of gravitational acceleration by simplifying the differential equation and error model of the inertial solution, and utilizing the coupling characteristics between inertial sensor error and navigation parameter error; then, according to the character that the change of velocity and position could be nearly zero under the condition of IMU in situ rotation, the pseudo observation was designed combing with the velocity and position of inertial solution, and the Kalman filter was established to estimate the sensor error; finally, calibration experiments of turntable mode and handhold mode were carried out. Experimental result showed that the bias and scale factors of MEMS accelerometers and gyros could be con verged within one minute with effective rotations of the held IMU.