论文部分内容阅读
针对多颗在轨卫星对空间合作目标的协同导航问题,提出一种适用于协同导航的分布式球面单形-径向容积求积分卡尔曼滤波(DSSRCQKF)算法。为了计算非线性滤波中的高斯加权积分,分别使用球面单形准则和二阶高斯-拉盖尔求积分准则计算球面积分和径向积分,提出了一种新的球面单形-径向容积求积分准则。将该准则嵌入分布式卡尔曼滤波框架中,结合协同导航的非线性数学模型,给出适用于协同导航的DSSRCQKF算法,该算法要求每颗导航星仅与其邻居星进行通信,通过数据的分布式融合实现对目标星轨道状态的一致估计,从而避免了传统集中式处理中较高的通信和计算压力。仿真实验结果表明,与分布式扩展卡尔曼滤波相比,本文算法将对合作目标的实时定位精度提高了11 m,定速精度提高了0.02 m/s,从而验证了本文算法的有效性。
Aiming at the problem of coordinated navigation of spacecraft in spacecraft orbiting on multiple satellites, a distributed spherical single-radial quadrature integrated Kalman filter (DSSRCQKF) algorithm is proposed for coordinated navigation. In order to calculate the Gaussian weighted integral in the nonlinear filtering, the spherical integral and the radial integral are calculated by using the single spherical form criterion and the second Gaussian-Laguerre integral integral criterion, respectively. A new spherical single form-radial volume seeking Points criteria. The criterion is embedded in the distributed Kalman filter framework. Combined with the nonlinear mathematical model of cooperative navigation, DSSRCQKF algorithm is proposed for coordinated navigation. The algorithm requires that each navigation star only communicate with its neighbors, and the distributed data The fusion enables a consistent estimation of the orbital states of the target, thus avoiding the high communication and computational pressures of traditional centralized processing. The simulation results show that compared with distributed extended Kalman filter, the proposed algorithm improves the real-time positioning accuracy of the cooperative target by 11 m and the fixed speed accuracy by 0.02 m / s, which verifies the effectiveness of the proposed algorithm.