Decentralized event-triggered control of large-scale systems with saturated actuators

Abstract

We consider a large-scale LTI system with multiple local communication networks connecting sensors, controllers, and actuators. The local networks operate asynchronously and independently of one another. The main novelty is that the decentralized controllers are subject to saturation. Our objective is to achieve a regional exponential stability providing a decentralized bound on the domain of attraction for each plant. We introduce a sampled-data event-triggering mechanism from sensors to controllers to reduce the amount of transmitted signals. Using the time-delay approach to networked control systems and appropriate Lyapunov-Krasovskii functionals, we derive linear matrix inequalities that allow to find the decentralized bounds on the domains of attraction for each plant. Numerical example of coupled cart-pendulums illustrates the efficiency of the method.