基于SRCKF算法的植保无人机定位估计

Location estimation on plant protection UAV based on SRCKF algorithm

基于高山丘陵地区地貌的复杂性、梯田间人工施药困难的背景下,针对植保无人机近地面喷洒任务的艰巨性,其自主导航定位技术的局限性,传统滤波估计算法对载体跟踪估计误差大、精度低、滤波不稳定等问题,提出了一种基于平方根容积卡尔曼滤波算法对植保无人机位置及姿态进行估计。较传统SRUKF、SRCDKF算法,文中算法既降低了状态量维数,又保证了数值计算效率及滤波精度。同等仿真环境下,实验结果表明,基于SRCKF-SLAM算法对载体位置估计精度最高,且稳定性最好;另外,在给定10°失准航偏角下,该算法对载体姿态角滤波估计,在近50 s内实现误差趋近于零,保证了植保无人机的精准导航。

Due to the complexity of landform in alpine and hilly areas and the difficulty of artificial application between terraces,plant protection UAV faces great challenge to undertake near-ground spraying tasks.Expect this problem,it has some limitation in autonomous navigation and positioning technology.Carrier tracking estimation conducted by traditional filter estimation algorithm shows great error,low accuracy and unstable filtering.In light of these problems,this study proposed a navigation algorithm of plant protection UAV based on square root cubature Kalman filter,which,comparing with the traditional SRUKF and SRCDKF algorithms,not only reduces the dimension of state quantity,but also ensures the efficiency of numerical calculation and the accuracy of filtering.Under the same simulation environment,the experimental results indicated that SRCKF-SLAM algorithm had the highest accuracy and the best stability in estimating the carrier position.In addition,given 10°misalignment yaw angle,the algorithm could filter the attitude angle of the carrier and realize the error approaching zero in nearly 50s,which ensures the accurate navigation of the plant protection UAV.