基于MAEKF算法的航姿参考系统设计

构建了以低成本MEMS陀螺仪、加速度计和磁传感器组合的航姿参考系统,提出了一个乘性自适应扩展卡尔曼滤波算法.取乘性误差四元数和陀螺仪误差作为状态量,基于重力场和磁场构造了量测矢量,用于修正航姿数据.并采用准确量测法,给滤波器加入了四元数的归一化约束,最后给出了基于新息的估计量测噪声方差矩阵的公式.通过仿真和试飞验证,表明本文设计的低成本的航姿参考系统能够提供比较准确的航姿信息.与常规的扩展卡尔曼滤波器比较,本文设计的乘性自适应扩展卡尔曼滤波算法有效提高了系统的精度和稳定性,并且具有较好的鲁棒性.

The Design of AHRS Based on MAEKF

An attitude and heading reference system(AHRS)is designed which is composed of low cost MEMS gyroscopes,accelerometers and magnetometer and a multiplicative adaptive extended kalman filter algorithm(MAEKF)is proposed based on the gravity field and magnetic field constructed measurement vector.In order to correct attitude and heading data,the MAEKF algorithm selects quaternion error and bias of gyroscopes as state vector.Then,the accurate measurement method is adopted to provide the filterwith normalized constraint.Furthermore,this paper presents an innovation-based formula which is used to estimate measurement noise covariance matrix.Finally,simulation and flight-test show that the designed AHRS can provide accurate attitude and heading data.Compared with traditional extended kalman filter(EKF),the MAEKF algorithm effectively improves the accuracy,stability and robustness of the system.