基于最小二乘法MEMS惯性传感器姿态解算算法

Attitude solution algorithm of MEMS inertial sensor based on least square method

针对MEMS惯性传感器在测量时出现误差不稳定和外界磁场变化所造成的姿态角误差较大问题,提出一种基于最小二乘法MEMS惯性传感器姿态解算算法.采用静态六位置法对三轴加速度计标定,对三轴加速度计建立误差模型,利用最小二乘法确定误差参数.根据欧拉角法中倾斜角和航向角分开求解特点,减小磁场变化干扰,再对欧拉角中存在的奇异值问题进行改进,分为一般姿态值和奇异值状态姿态值两种滤波模式.实验结果证明,通过对三轴加速计的标定补偿和欧拉角的奇异值问题改进后得到的姿态角精度高,并且在遇到磁场变化时稳定性优于四元数法,不会出现奇异值.

Aiming at the large attitude angle error caused by error instability and external magnetic field variation during measurement of MEMS inertial sensor,an attitude solution algorithm based on least square method was proposed.The static six-position method was used to calibrate the three-axis accelerometer,the error model was established for the three-axis accelerometer,and the least square method was used to determine the error parameters.According to the characteristic of solving tilt angle and heading angle separately in euler angle method,the interference of magnetic field variation was reduced,and the singular value problem in euler angle was improved,which was divided into two filtering modes:general attitude value and singular value state attitude value.The experimental results showed that the improved attitude angle obtained by the calibration and compensation of the triaxial accelerometer and the singular value problem of the euler angle has high precision,and was more stable than the quaternion method when the magnetic field changed,and did not appear singular value.