Event-triggered adaptive control of uncertain non-linear systems under input delay and limited resources

Abstract

In this paper, both input delay and limited bandwidth challenges in uncertain networked control systems have been tackled. For this purpose, an event-triggered adaptive controller based on backstepping technique is designed for a general class of non-linear uncertain MIMO systems. The event-triggered mechanism is placed in the sensor-to-controller channel to reduce communication burden and save computational resources. Rather than predefined fixed or simple relative triggering thresholds, the proposed triggering condition is derived based on the negative semi-definiteness of derivative of Lyapunov function. To handle input delay problem, an auxiliary compensation system based on integral of input signal is introduced in the controller design. System stability are guaranteed under proposed scheme with no Zeno behavior. Simulation results on three-link cylindrical robotic arm show that the proposed control scheme successfully compensates for input delays and ensures a substantial saving in computational and network resources characterized by number of required control updates and signal transmissions over the network.