基于SE(3)-EKF的旋翼无人机GNSS/SINS空中快速对准方法

Rapid in-flight alignment method for rotor UAV with GNSS/SINS based on SE(3)-EKF

针对旋翼无人机GNSS/SINS组合导航系统采用传统对准方法难以在大失准角下实现空中快速对准的问题,提出了一种基于SE(3)-EKF的旋翼无人机GNSS/SINS空中快速对准方法。将姿态和速度状态误差量构造为李群元素,并考虑陀螺仪和加速度计零偏误差,构建了群-矢量混合误差模型,推导了基于SE(3)-EKF的GNSS速度量测方程。实验结果表明:当失准角为[3°, 4°, 95°]时,所提方法在200 s内实现了约为0.1°的水平姿态对准精度,较传统的扩展卡尔曼滤波(EKF)、无迹卡尔曼滤波(UKF)分别提高了50%以上和20%以上;航向角误差约为0.4°,相较于EKF、UKF分别降低了45%以上和43%以上。当失准角为[20°, 20°, 175°]时,EKF和UKF已无法收敛,而所提方法水平姿态误差可收敛到约0.1°,方位误差可收敛到约0.5°,验证了所提方法适用于不同场景下的初始失准角,具备一定的工程应用价值。

Aiming to address the challenge of achieving rapid airborne alignment under large misalignment angles with traditional alignment methods for rotary unmanned aerial vehicle(UAV)GNSS/SINS combined navigation systems,a rapid in-flight alignment method for rotor UAV with GNSS/SINS based on SE(3)-EKF is proposed.The errors in attitude and velocity states are formulated as elements of a Lie group,and a group-vector hybrid error model is constructed considering gyroscope and accelerometer bias errors.The GNSS velocity measurement equation based on SE(3)-EKF is derived.The experimental results show that when the misalignment angles are[3°,4°,95°],the proposed method achieves horizontal attitude alignment accuracy of about 0.1°within 200 s,which is more than 50%higher than the traditional extended Kalman filter(EKF)and more than 20%higher than the unscented Kalman filter(UKF).The azimuth misalignment error is approximately 0.4°,which is more than 45%lower than the EKF and more than 43%lower than the UKF.When the misalignment angles are[20°,20°,175°],the EKF and UKF fail to converge,while the horizontal attitude errors of the proposed method can converge to about 0.1°,and the azimuth error can converge to about 0.5°,which proves that the proposed method is suitable for the initial misalignment angles in different scenarios and has certain engineering application value.