基于模糊卡尔曼滤波的四旋翼姿态估计方法

A QUADROTOR ATTITUDE ESTIMATION METHOD BASED ON FUZZY KALMAN FILTER

针对四旋翼在飞行中因运动加速度及磁场干扰等影响引起的姿态估计误差,提出一种基于动态估计的模糊扩展卡尔曼滤波姿态估计算法。该算法利用传感器数据对运动频率、运动加速度及磁场干扰的大小进行估计,对估计值模糊推理后,使用输出值调整卡尔曼滤波测量噪声方差和梯度下降步长,增加姿态估计算法的动态适应性。为验证所设计算法的估计效果,使用四旋翼仿真模型进行验证。实验结果表明,该算法能够有效地提高四旋翼在高机动、强干扰情况下姿态估计的姿态跟踪精度、动态性能、抗干扰能力。

Aimed at the attitude estimation error caused by the acceleration and magnetic disturbance of quadrotor in flight,a fuzzy extended Kalman filter attitude estimation algorithm based on dynamic estimation is proposed.The sensor data were used to estimate the motion frequency,acceleration and magnetic field interference.After the fuzzy inference of the estimated values,the Kalman filter was used to adjust the measurement noise variance and gradient step size to increase the dynamic adaptability of the attitude estimation algorithm.In order to verify the estimation effect of the designed algorithm,a quadrotor simulation model was used to verify the estimation effect.The results show that the algorithm can effectively improve the attitude tracking accuracy,dynamic performance and anti-jamming ability of quadrotor attitude estimation under high maneuvering and strong interference.