Optimal Disturbance Rejection Methods For Systems With Time-delays In High-speed Networks

In this paper, the problem of optimal disturbance rejection for systems with time-delays in high-speed networks is considered. The main contents are given as follows:1. The mathematical models for systems with delayed control and measurement in high-speed networks are established and then the discrete-time models are founded.2. Based on a model transformation, the discrete-time system with control and measurement delays is transformed to a formal nondelay system. Then the optimal control law with a state feedback, a disturbance feedforward and a control memory term is derived from a discrete Riccati equation and a Stein equation. The feedforward control term and control memory term compensate for the effects of disturbances and the control delay, respectively. A disturbance state observer is constructed to make the feedforward compensator physically realize. Simulation results demonstrate the effectiveness of the optimal control law.3. Consider the optimal disturbance rejection for nonlinear discrete-time systems with delayed control and measurement in high-speed communication networks via a high-gain observer. Based on a model transformation, the optimal control law via the high-gain observer is constructed and the high-gain observer, which can remove the uncertainty come from disturbance model transformation, is proved. By linearizing the system around the desired equilibrium point, a state feedback controller is designed for the linearization system under disturbances. The Lyapunov function for the closed-loop system in the neighborhood of the origin is used to estimate the region of attraction. The stability of this system is proved.4. Consider the optimal disturbance rejection for linear discrete-time systems with delayed control and measurement in high-speed communication networks, by constructing the internal model principle, make the better control signal to overcome disturbance signal, design the stabilizing controller can stabilize the closed-loop system to ensure the stability of the closed-loop system. Use the internal model principle, by designing the internal model controller with a dynamic model compensator, make the unsteady-state error, complete disturbance rejection of the system.