The detection and identification of Resident Space Objects (RSOs) from survey tracks requires robust and efficient orbit determination methods for the association of observations of the same RSO. Both Initial Orbit Determination (IOD) and Orbit Determination (OD) methods perform the orbital estimation in which the association of tracks relies. The choice of proper IOD and OD methods is essential for the whole data association, since they are in charge of providing the estimation required to evaluate the figure of merit of the association. In this paper, we review the state of the art and propose a novel method that does not require initialisation, accounts for measurement noise and provides a full estimation (i.e., state vector and covariance) from an arbitrary number of optical observations. To do so, a boundary value problem is formulated to find a pair of ranges leading to a minimum residuals of the observations. The proposed methods are compared against classical alternatives simulated in scenarios representative of the current space debris environment.

从勘测轨迹中检测和识别驻地空间物体 (RSO) 需要稳健且高效的轨道确定方法来关联同一 RSO 的观测。初始轨道确定 (IOD) 和轨道确定 (OD) 方法都执行轨道关联所依赖的轨道估计。选择合适的 IOD 和 OD 方法对于整个数据关联至关重要，因为它们负责提供评估关联品质因数所需的估计。在本文中，我们回顾了现有技术并提出了一种不需要初始化、考虑测量噪声并从任意数量的光学观测值提供完整估计（即状态向量和协方差）的新方法。为此，边界值问题被公式化以找到导致最小观测残差的一对范围。将所提出的方法与在代表当前空间碎片环境的情景中模拟的经典替代方法进行比较。