Stabilizing a Nonlinear Model-based Networked Control System with Communication Constraints

In this paper, a class of nonlinear networked control systems (NCS) operating over a shared-channel are considered. For sensor networks and networks requiring information collaboration among various devices, the objective to reduce contention and use the bandwidth limited network resources more efficiently can be achieved by developing "smart" sensors. Model-based control and event-based triggering can be fused such that smart sensors determine the "value of information" for stable operation of control systems. A context-aware feedback policy can be developed for a class of nonlinear NCS based on the informational value of sensor measurements that minimizes network usage or traffic. The developed aperiodic feedback policy guarantees global asymptotic tracking of output states of an uncertain system along the desired time-varying trajectory. A direct adaptive parameter update law is formulated to estimate the uncertain system dynamics that can further reduce feedback requirements. A piecewise continuous tracking controller is developed and validated using extensive simulation results for nonlinear scalar and coupled MIMO systems.