Search

This paper studies the pose tracking control problem for satellite proximity operations between a target and a chaser satellite, by which we mean that the chaser is required to track a desired time-varying trajectory given in advance with respect to the target. Firstly, by consulting an adaptive sliding-mode control method in literature developed for a class of nonlinear uncertain systems, an effective pose tracking controller is obtained. This controller requires no information about the mass and inertia matrix of the chaser, and takes into account the gravitational acceleration, the gravity-gradient torque, the J 2 perturbing acceleration, and unknown bounded disturbance forces and torques. Then, an updated controller is proposed by combining the aforementioned controller and the unscented Kalman filter (UKF). This updated controller estimates the inertial parameters of the chaser through UKF, so it is of better adaptive ability to the initial estimation of the inertial parameters. Finally, numerical simulations are given to demonstrate the effectiveness of the proposed controllers. The simulation results show that the updated controller is more accurate. • An adaptive sliding-mode (ASM) controller is designed for pose tracking, and both unknown bounded disturbances and the uncertainty of the inertial property are considered. • The adaptive ability of the proposed method to the initial estimation of the inertial parameters is improved by combining ASM and UKF. • Our methods have a wider application and need less computing resource. [ABSTRACT FROM AUTHOR]

Abstract Capturing an uncontrolled space target is a tremendously challenging research topic. Target capture by a space robot can be well planned according to predicted motion of the target. In this paper, motion prediction of an uncontrolled space target is studied and a motion prediction algorithm is proposed. In the proposed algorithm, firstly a method for identifying the parameters of motion state and inertial property of the target is established; and then through substituting the identified parameters into the dynamic equations of the target, the motion of the target can be predicted as the solution of the equations. In the identification of the parameters, the unscented Kalman filter (UKF) is applied. In order to support the UKF, a method for estimating noise level of the observation data is developed, so our motion prediction algorithm is noise adaptive. A practical convergent criterion is also designed to determine the time when the estimated result of the UKF is accurate enough, such that the predicted motion is credible enough. After that, the accuracy of the prediction is further improved by an optimization method. In the end of this paper, numerical simulations are done to verify the validity of the proposed motion prediction algorithm. Simulation results indicate that the proposed algorithm is able to predict the motion of the target precisely. [ABSTRACT FROM AUTHOR]