Maneuver detection and tracking of a space target based on a joint filter model

Orbit maneuver of non‐cooperative space targets is difficult to detect, and thus greatly affects the accuracy of real‐time target tracking. To solve this problem, a joint filter‐based algorithm is proposed to detect and track targets with random pulse maneuvering. In the algorithm, the joint filter is constructed using two nonlinear filters. The first is an auxiliary filter that utilizes the kinematic model to predict the target motion and estimate the target movement according to the continuity of Keplerian elements. The other is a primary filter that uses an orbit dynamic model to predict the target motion and utilize the estimated results provided as feedback by the auxiliary filter as the constraints to re‐estimate the target orbit in real time. Moreover, simulations were conducted to detect and track non‐cooperative space targets with a series of random pulse maneuvers, and thus validate the efficiency of the new algorithm.