Fuzzy-model-based control for nonlinear networked systems with random packet losses

This paper investigates the problem of observer-based controller design for nonlinear networked control systems subject to imperfect communication links. The Takagi-Sugeno (T-S) model is utilized to approximate the nonlinear plant. Both the imperfect sensor-to-controller and controller-to-actuator links are considered, which are described via two independent Bernoulli stochastic processes. A dynamic compensation strategy is adopted in the controller-to-actuator link. The premise variables of the observer and the controller to be designed depend on the state variables estimated by the observer instead of those of the plant. Moreover, sufficient criteria are obtained to guarantee the resulting closed-loop system to be stochastically stable with H_∞ performance. Finally, a numerical example is provided to illustrate the effectiveness of the methodology proposed in this paper.