Synchronous Control Of Two Types Of Time-delay Recurrent Neural Networks Based On Event-triggered

Event-triggered control is a new control strategy that will perform control tasks only when the stability and performance of the system are required,and has a wide range of development space.Compared with the traditional periodic control strategy,event-triggered control can effectively reduce the number of control tasks and save network resources under the premise of ensuring system performance.Although event-triggered control has a good application prospect,there is still a lack of relevant mathematical theoretical results in the field of recurrent neural network control system,leading to most of the recurrent neural network control systems are still using periodic control strategies,which wastes network resources.Therefore,under the theoretical framework of event-triggered control,this paper mainly studies the following two aspects of recurrent neural network control system:(1)Based on the event-triggered mechanism,the synchronous control problem of continuous-time recurrent neural network with time-varying delay is studied.Two systems are considered,one is a recurrent neural network driven system;The second is the recurrent neural network response system.According to the Lyapunov stability theory,a suitable Lyapunov-Krasovskii functional is constructed,a feasible event-triggered mechanism is designed,the parameters of the event-triggered mechanism are set,and sufficient conditions for driven-response system synchronization are proposed,which effectively reduces the amount of computation and the number of controller updates,and reduces the network load.In addition,the feasible solution of the system stability is obtained by linear inequality technology,so that the effectiveness of the set controller is verified by numerical simulation.(2)On the basis of the previous article,the synchronous control problem of discrete-time recurrent neural network with time-varying delay based on eventtriggered is further studied.Firstly,a discrete-time recurrent neural network system is used as the driven system to construct a response system with the same structure as the driven system and different initial values.By applying a suitable event-triggered controller to the driven-response system,the driven system and the response system are synchronized asymptically.Then,using Lyapunov stability theory,the Lyapunov-Krasovskii functional was constructed,and the sufficient conditions for system synchronization were derived by introducing mathematical methods such as free weight matrix and Jensen’s inequality.Finally,multiple numerical simulations verify the feasibility of the theory and the effectiveness of the set controller.