一种自适应互补滤波姿态估计算法

An Adaptive Complementary Filter for Attitude Estimation

针对微型四旋翼飞行器捷联惯性导航姿态解算中低成本MEMS器件易发散导致姿态漂移问题,基于四元数姿态更新算法,提出了一种自适应互补滤波姿态估计算法,该算法利用MEMS器件中的加速度计补偿陀螺仪偏差引起的姿态漂移误差,并根据陀螺仪频率响应的截止角速度分段调整补偿系数。最后,基于ARM Cortex-M3内核设计了微型四旋翼飞行器,实验结果表明,该自适应互补滤波算法能有效、实时地跟踪四旋翼飞行器的三维姿态,提高姿态解算的测量精度,实现了飞行器的自主悬停。

In order to overcome the drift problem of low-cost MEMS inertial sensor in SINS attitude estimation, this paper presents a quaternion-based adaptive complementary filter for attitude estimation. The proposed algorithm uses the MEMS accelerometer to eliminate attitude estimation error caused by gyroscope bias and drift, and the compensation coefficient is adjusted by the piecewise function defined by gyroscope frequency response of cut-off angular velocity. Finally, a micro quadrotor is designed based on ARM Cortex-M3. The experimental results show that the adaptive complementary filter is able to effectively trace the quadrotor’s attitude in real time, improve the attitude estimation accuracy and then implement the autonomous hover of a micro quadrotor.