Jagannathan Neural Network Approximation-based Event-Triggered Control of Uncertain MIMO Nonlinear Discrete Time Systems

This paper proposes neural network (NN) approximation-based event-triggered control of multiple-input and multiple output (MIMO) nonlinear discrete-time systems in the context of limited communication over the network. Unlike the traditional NN-based discrete-time control, the weights are updated non-periodically and only at the trigger instants. The Lyapunov direct approach is utilized to arrive at an analytical condition referred to as event-trigger condition which guarantees the uniform ultimate boundedness (UUB) of the system states and NN weight estimation errors. This design not only reduces the network communication but also the computation due to non-periodic control execution and NN weight update. In addition, explicit knowledge of the system dynamics is not necessary for generating the control input. Finally simulation results corroborate the analytical claims.