ABSTRACT
In this chapter, the design process is revisited so that robustness can be incorporated. The procedure (Network Control Systems (NCS) modeling and communication sequence optimization) remains the same but the cost function is reformulated. A quadratic cost function was constructed and a cost was given by solving a Lyapunov function. The cost function is a measure of 'how good' a particular schedule is, i.e. it maps a communication sequence to a scalar. An optimal sequence is the one that minimizes that scalar. A robust design is not optimal for just the given mathematical model, but it is optimal for a whole family of possible models. The NCS model used here is the one of a generalized plant in feedback with a two Degrees Of Freedom controller. The network is modeled as schedulers.
