Neural network-based adaptive event-triggered control of nonlinear continuous-time systems

In this paper, a neural network (NN) based adaptive event-triggered control is developed for a single input and single output (SISO) uncertain nonlinear continuous time system. An explicit design of the event-triggered controller using NN approximation and feedback linearization is presented. The controller dynamics are approximated by using two single layer NNs. In addition, novel weight update laws are derived for the NNs in the context of event-triggered transmission, i.e., weights are updated only at the triggering instants, hence, aperiodic in nature. The closed loop stability analysis using Lyapunov approach for impulsive dynamical system is carried out to show the uniform ultimate boundedness (UUB) of the NN weight estimation errors as well as system states. Numerical results are included for validating the design.