This research investigates an integrated predictor-observer feedback control strategy for the vibration suppression of large-scale spacecraft affected by unbounded input time-delay effect. The vibration model incorporates orbit-attitude coupling effects, with lumped disturbance introduced to address nonlinearities. Resulted from large-scale characteristics, the unbounded input time-delay is considered during the dynamic response process. Accordingly, for the infinite property of delay effect in time domain, a universal integrated predictor-observer feedback control scheme is developed to achieve asymptotic stabilization of closed-loop system via Lyapunov theory. Notably, the state, disturbance and intermediate observers are analyzed subject to unbounded time-delay respectively to estimate both states and disturbance simultaneously, in contrast to the availability or ignorance assumption directly in existing literature. Then a predictor is introduced to compensate for the time-delay and prevent adverse performance impacts. Furthermore, numerical simulations of the spacecraft's vibration dynamics are performed to validate the effectiveness of proposed control strategy.

This paper presents a novel dynamic scaling–based non-certainty-equivalent adaptive control method for the attitude tracking control problem of spacecraft in the presence of inertia uncertainties. By virtue of the modified dynamic scaling factor and an angular velocity observer, the proposed controller provides a more direct approach to circumvent the integrability obstacle of the basic immersion and invariance (I&I) adaptive method without any information about the bounds of system uncertainties. Compared with the existing filter-based non-certainty-equivalent adaptive attitude controller, the proposed controller achieves significant order reduction in closed-loop systems and removes the constraint on initial values of the parameter lock property. The dynamic gains in the proposed method are modified to have linear growth in the dynamic scaling factor and restricted by the given upper bound in order to avoid bad transient performance. Numerical simulations are presented to highlight the superiority of the proposed method compared with the filter-based I&I control method.