Predictor-based observer and resilient controller design for aperiodic sampled-data systems with disturbance and output delay

In this paper, the control problem based on state/disturbance observers is studied for a class of networked aperiodic sampled-data systems with unknown disturbances and output delays. A novel observer structure is devised to estimate states and disturbances by predicting the actual output of the system. Moreover, the disturbance constraint is broadened to encompass wider types such as unbounded finite derivatives. Based on the obtained estimation of disturbance and state, a resilient controller is proposed to compensate for the impact caused by controller parameter perturbations. In particular, a new class of Lyapunov-like functionals is constructed to extend the sampling interval associated with exponential convergence. By employing matrix analysis and integration techniques, sufficient criteria are established to guarantee the exponential convergence of the networked aperiodic sampled-data closed-loop dynamics. The obtained criteria reveal that the estimation error of the disturbance depends only on the errors of the predictor and state observation, benefiting from the novel structure of the devised observer. The parameter gains of the observer and controller are readily determined by solving a set of convex optimisation constraints. The effectiveness and superiority of the proposed observer-based control algorithm are confirmed through developed examples.