Various spacecraft missions have driven the need for lighter, stronger deployable structures, which help to hold instruments, such as a magnetometer, away from the spacecraft to avoid the disturbance caused by remanence of the spacecraft body. In this paper, we will present a type of deployable manipulator for small spacecraft, which is characterized by a small stowed volume, light weight and a large magnification ratio. Because the actual parameters of the deployable manipulator are all nonlinear, modeling of the manipulator becomes a key point. Considering the uncertainties in the model parameters, an approximate dynamic model with uncertain parameters is formulated by classifying the uncertainties into different types, including constant parametric, variable parametric and nonparametric uncertainties. Then, a robust adaptive control strategy is proposed to compensate for or reject these uncertainties separately; an output feedback controller is designed to reduce the errors between the desired and the real trajectories, an adaptive controller aims at compensating for constant parametric uncertainties and a robust controller is used to reject the variable parametric and nonparametric model uncertainties. Thus, a robust adaptive control strategy does not rely on the exact dynamic model and can completely compensate for or reject the effect of model uncertainties. Finally, the simulation results show that the proposed control law is perfectly adequate for the deployable manipulator.

This paper is concerned with robust tracking control of electrical driven free floating space robot manipulators with unmodeled uncertainties and external disturbances. The dynamics of the free-floating space manipulator system derived using virtual manipulator approach (VMA) uses only link angles as generalized coordinates, which is suitable for the controller design in joint space. Since joint coupling, unmodeled uncertainties, as well as external disturbances are treated as lumped disturbance, then a proportional plus derivative (PD) feedback control with disturbance observer (DOB) was developed and implemented at each joint of space manipulators including electrical and mechanical dynamics of motor. In addition, a feed-forward control consisting a simple predictor was added to reduce tracking error further. Simulation results from a six-link electrical driven space manipulators show that the developed controller achieves superior performance with less tracking error, especially when internal model uncertainties and large external disturbances are present