Dynamic Event-triggered Synchronous Control Of Half-horse Jumping Neural Networks

Semi-Markov jumping neural networks have attracted much attention because it can better model and describe the physical system,including stability analysis and control,and has been widely used in vertical take-off and landing helicopters,DNA analysis and DC motor control.Many important researches on semi-Markov jump neural networks have been reported,such as stability analysis,extended non-fragile dissipation estimation and control design.On the other hand,unlike the past point-to-point system,the current network control systems are implemented by network-connected components.Time-triggered control executed in a periodic manner will lead to increased computational complexity,waste of communication resources and inefficient use of limited network resources.On the other hand,according to the selected threshold,the use of the event triggers strategy only allows data to be sent when the event trigger condition is met.Based on the dynamic event trigger scheme,this paper studies the synchronization problem of the semi-Markov jump neural networks.The main contents are as follows:1)The finite time synchronization problem of a semi-Markov neural networks with timevarying delay are studied.The main purpose is to use the dynamic event trigger scheme to synchronize the semi-Markov neural networks in a finite time and reduce the number of sampling points.According to the Lyapunov stability theory,a mode-dependent Lyapunov function is constructed.Compared with the traditional static event triggering scheme,the dynamic event triggering scheme is used to adjust the amount of data transmission and reduce the burden on the network.The general free weight matrix is used to estimate the single integral term,a less conservative conclusion is obtained in the standard linear matrix inequality.Finally,the superiority of the method is verified by simulation results.2)The synchronization problem of a stochastic semi-Markov jump neural networks with timevarying delays are studied.The main idea is to use a stochastic differential equation with semiMarkov jump parameters to describe the neural networks.First,we use supplementary variables technology and system transformation to transform the finite semi-Markov process into a related Markov process.Secondly,through the stochastic analysis method and the La Salle-type invariance principle,a sufficient condition for the synchronization of the semi-Markov jump neural networks is proposed,and compared with the existing methods,the results are less conservative.Finally,the application of the industrial four-tank model verifies the effectiveness of the algorithm.3)It solves the synchronization problem of stochastic semi-Markov jump neural networks with general incomplete transition rate.The main goal is to consider the incomplete transition rate problem based on stochastic semi-Markov process,stochastic analysis theory and state feedback control technology.First of all,applying the extended Jensen integral inequality and Wirtinger inequality,a new synchronization criterion is obtained.Secondly,the upper bound of the variable sampling interval and the sawtooth structure information of the variable input delay are fully utilized,and the required dynamic event trigger controller is designed.In addition,using the Lyapunv-Krasovskii functional method and the La Salle-type invariance principle,new sufficient conditions for the synchronization of the semi-Markov jump neural networks are established.Finally,the single-link manipulator model is used to verify superiority of the main results.