Event-triggered state tracking for two-dimensional neural networks with impulsive learning control schemes

Abstract

In this paper, different types of learning control schemes are proposed to study the tracking of two-dimensional discrete neural networks. The learning control schemes combine the advantages of impulsive control and iterative learning control strategies, because the impulsive control technique can improve tracking performance rapidly. Further, the event-triggered mechanism is used to determine the impulse time. And an equivalent system is proposed by constructing a trigger function, which is used to get over the difficulties in the theoretical analysis. Then learning control schemes are designed in line with the equivalent system, and some sufficient conditions are proposed to guarantee the convergence of the tracking error. The main results show that the tracking performance can be improved effectively by our control schemes. And it shows that our control schemes are more effective than traditional learning control approaches. Finally, the effectiveness is illustrated by numerical simulations.