基于自适应互补滤波的四旋翼飞行器姿态解算

Attitude Algorithm of Quadrotor Based on Adaptive Complementary Filtering

在四旋翼无人机中,姿态传感器采用捷联式惯导惯性检测单元(IMU),其中包括加速度计、陀螺仪、电子罗盘和空气压力高度计。这些传感器在工作过程中存在温度漂移以及噪声干扰,为了得到准确的姿态数据,首先建立了传感器四元数模型,在频域中设计互补滤波器,并设计了PI自适应补偿系数,对传感器数据进行融合、补偿和修正,有效地避免了系统模型误差对姿态估计的影响。修正后的角速度通过一阶龙格-库塔法、四元数算法完成飞行器的姿态解算。传统互补滤波器在噪声大时滤波效果不理想,故加入PI控制,形成一种效果更好的自适应滤波算法,根据仿真结果,该算法滤波后的信号比传统互补滤波的结果更加平滑,更接近理想波形。

The strapdown inertial navigation inertial measurement unit（IMU） is used as attitude sensor in quadrotor, which includes accelerometer, gyroscope, electronic compass and barometer. These sensors have temperature drift and noise disturbance during operation. In order to obtain accurate attitude data, a quaternion model of the sensor was established firstly, and a complementary filter and PI adaptive compensation factors were designed in frequency domain, and the sensors＇ data were coalesced, compensated and corrected, which effectively avoid the impacts of system model error on attitude estimating. The corrected angular velocity was used to achieve the attitude calculation of the aircraft through the Runge-Kutta method and the quaternion algorithm.The traditional complementary filters are not ideal for filtering when the noise is strong, a more effective adaptive filtering algorithm is designed when adding the PI control. According to the simulation result, the filtered signal of the algorithm is smoother than the traditional complementary filtering signal and is closer to the ideal waveform.