Adaptive Tracking Method for Non-Cooperative Continuously Thrusting Spacecraft

Abstract

Performance of the traditional Kalman filter and its variants can seriously degrade when they are used to track a non-cooperative continuously thrusting spacecraft. To overcome this shortcoming, an adaptive tracking method for relative state estimation of a non-cooperative target is proposed based on the interacting multiple model (IMM) algorithm. First, built upon a current statistical jerk (CSJerk) model, a robust CSJerk filtering (RCSJF) algorithm is developed, which can address the issue of low estimation accuracy and instability of traditional approaches at the moments when the spacecraft starts and ends thrusting. Second, the developed RCSJF algorithm is further used to form the model set of the IMM by incorporating different maximum jerk values, based on which an adaptive tracking method is presented that can track a non-cooperative target with different maneuvering levels. Simulation results show that the proposed method can effectively track the target across all thrusts levels under the conditions considered, and the convergence performance of the proposed method is improved in comparison to the CSJerk-based extended Kalman filter, especially at the start and end time of the maneuver.