基于四元数改进型互补滤波的MEMS姿态解算

Attitude estimation of MEMS based on improved quaternion complementary filter

针对四旋翼飞行器的姿态解算问题,设计了由STM32、陀螺仪、加速度计等组成的姿态测量单元,采用四元数法对姿态角进行描述,详细分析了姿态解算的过程,建立了系统及传感器模型,并在互补滤波算法的基础上,构造了一种改进型互补滤波器来完成姿态解算。实验测试表明,与传统的卡尔曼滤波算法和互补滤波算法比较,改进型互补滤波算法的数据平滑能力更强,收敛速度更快,有效地实现了飞行器姿态数据的融合,提高飞行器的姿态估计精度,满足了小型四旋翼飞行器姿态控制的要求,增强了系统的鲁棒性与稳定性。同时通过实际飞行实验,验证了算法的可行性。

Aiming at the attitude estimation of quad-rotor helicopter, the attitude measurement unit was designed, which consists of STM32, gyroscope, and accelerometer etc. The quaternion algorithm was used to describe attitude angles. The process of attitude estimation was analyzed in detail, and the models of the sensor and system were es- tablished. Furthermore, an improved complementary filter to estimate the attitude was designed based on the com- plementary filtering algorithm. The experimental results indicate that, comparing with the traditional Kalman algo- rithm and complementary filtering algorithm, the improved complementary filter effectively implements the attitude＇s data fusion of aircraft and meets the attitude＇ s controlling needs of miniature quad-rotor helicopter. At the same time, the improved algorithm has stronger data smoothing ability and faster convergence rate.