Command Filtered Event-Triggered Adaptive Control for a Class of MIMO Nonlinear Systems Based on Neural Network Model

This article deals with the tracking control problem for a class of multi-input and multioutput (MIMO) nonlinear systems with uncertain dynamics under the premise of feedback path transmitted by dynamic event-trigger mechanism. The neural network adaptive plant model is designed to generate predictive system states for controllers. Command filters are introduced to fix the jumping problem of virtual controllers while avoiding the issue of "explosion of complexity" caused by the recursive differentiate behavior in conventional event-triggered backstepping controllers design. Moreover, dynamic event-trigger conditions are constructed to decide the feedback path aperiodically transmit plant states instants. Simulation results indicate that this proposal can reduce the communication times considerably without degrading system performance.