Event-triggered adaptive control for nonlinear systems using time-receding horizons without initial dependence.

This paper investigates the full-state constraint event-triggered adaptive control for a class of uncertain strict-feedback systems. The lack of information on the coupling dynamics of virtual variables in backstepping increases the complexity of feedback design. Given this, the requirements of shaping system performance constraints, eliminating initial dependence, and reducing data transfer costs together give rise to an interesting and challenging problem. Constructing the time-receding horizon (TRH) and stitching it with the quadratic Lyapunov function (QLF) is the key to constrained tracking. Specifying TRHs as a set of smooth bounds with fixed-time convergence and forcing the system to stabilize within the constrained region before the prescribed settling time provide a sufficient condition for practical finite-time stability (PFS). For relaxing the initial dependence, a tuning function is designed to match the performance constraints under arbitrary system initial conditions. A dual-channel event-triggered mechanism (ETM) is developed to automatically adjust the controller and estimator data flow updates with less transmission burden. By combining a specific inequality with backstepping, uncertainties are overcome without the "complexity explosion" in recursion steps. Finally, simulations demonstrate the effectiveness of the proposed method. • The tracking issue of strict-feedback systems with full-state constraints is studied. • An approximation-free event-triggered adaptive control scheme is developed. • The tuning function removes the initial dependence inherent in constrained control. • Novel time-receding horizons for the practically finite-time stability are proposed. • The dual-channel event-triggered mechanism reduces the data-transmission burden. [ABSTRACT FROM AUTHOR]