论文部分内容阅读
如果实现机动的时刻,推力向量值和方向对完成跟踪宇宙飞船的观察者是未知的,则对应用推力进行机动的宇宙飞船的轨道参量进行跟踪是一个复杂的问题。普通的递推滤波方框图[1-3]认为存在有关所加的速度脉冲值和时刻的预先信息,因此,它们在一般情况下对解决此问题是不能用的。对被动轨道运动的宇宙飞船跟踪时,估算误差的协方差矩阵和卡尔曼滤波器的传递系数趋向于零,而辅助测量对估算有较小的影响[l-4]。当被估算的轨道参量突然变化和卡尔曼滤波器的传递系数快速趋向零时,运动参量的估算开始很快发散,并与真实值相差很大。
If moments of maneuvering, thrust vector values, and direction are unknown to the observer who completes the tracking of the spacecraft, tracking the orbital parameters of the powered spacecraft using thrust is a complex issue. Ordinary recursive filtering block diagrams [1-3] consider that there is advance information about the added speed pulse values and moments so they are generally not available to solve this problem. When tracking a orbiting spacecraft, the covariance matrix of the estimated error and the transfer coefficient of the Kalman filter tend to be zero, whereas the secondary measurement has less impact on the estimation [1-4]. When the estimated sudden changes in orbital parameters and the transfer coefficient of the Kalman filter tend to zero quickly, the estimation of the motion parameters begins to diverge rapidly and differs greatly from the true value.