Adaptive Event-Triggered Output-Feedback Stabilization With Exponential Convergence

This paper seeks adaptive event-triggered output-feedback control which enables exponential stabilization for nonlinear systems with unknown growth rate. Convergence rate, as an important performance specification, is usually hard to acquire in the context of adaptive control. Besides, the event-triggered architecture could undermine convergence rate, entailing a competent compensation mechanism under reduced execution. As such, we are compelled to pursue a distinctive adaptive event-triggered output-feedback scheme. Specifically, a delicate dynamic gain incorporating exponential-type time-varying information is introduced, which would not only counteract the unknown growth rate, but particularly enable desired convergence. Correspondingly, a compatible event-triggering mechanism, capable of ensuring timely execution for adaptive compensation, is designed by suitably exploiting the gain information. In this way, an adaptive event-triggered output-feedback controller is constructed, which can render exponential convergence for system states, alongside an explicit pre-estimate for inter-execution intervals
Note to Practitioners
—In the networked setup, the communication resources are usually shared but limited, entailing efficient resource utilization. Notably, event-triggered control could cater for the efficiency demand, which requires information transmission and control updating only when necessary for systems, unlike sampled-data control with a conservative prescribed rate of sampling/execution. This paper is devoted to developing a distinctive adaptive output-feedback scheme of event-triggered stabilization under the architecture with controller-to-actuator communication reduced. The scheme can not only cope with system nonlinearities with unknown growth rate (resulting from e.g., slowly-but-widely varying parameters and multiplicative disturbances) but particularly enable exponential convergence rate, an objective which is impossible for the existing schemes. Besides, the scheme allows flexible parameter choice, with which the change in the number of executions is illustrated through simulation.