State Estimator And Filter Design With An Adaptive Event-triggered Communication Scheme For A Class Of Neural Networks

Neural networks play an important role in the research and development of pattern recognition,signal processing,optimizational calculation as well as associative memory.With the rapid development of these research areas,there will be a higher requirement for the study of neural networks.In order to carry out further study of neural networks,we must obtain the internal states of neural networks.Therefore,the research of state estimation of neural networks has important theoretical significance and practical application value.However,in real network environment,because of the limitation of network bandwidth,it's impossible to estimate all of the signals.Therefore,the event-triggered scheme is used in the research of network control to reduce the unnecessary waste of bandwidth resources.Although the event-triggered mechanism is efficient in reducing the burden of network bandwidth,it affects the stability of the system,which leads to the inaccuracy of state estimation.So,to achieve a certain degree of balance between the network bandwidth resources and the stability of the system,this paper considers the problem of state estimator and H_? filter design based on the adaptive event-triggered communication scheme.The main research contents of this paper are as follows:1.The second chapter is about the problem of neural network state estimator design based on the adaptive event-triggered mechanism.The rationality of the state estimator is proved by Matlab/Simulink simulation.Firstly,a state estimator of neural networks based on the adaptive event-triggered mechanism is designed;Next,the asymptotic stability of the system is proved by constructing an appropriate Lyapunov functional;Then,the gain of state estimator is obtained by lingering the matrix inequality;Finally,a numerical example is presented to demonstrate the rationality of the state estimator and the advantages of the adaptive event-triggered mechanism are proved.In this chapter,the state estimator is highly related to the parameters of the system,but the form of the state estimator in this chapter is relatively simple.2.In the third chapter,because of the low accuracy of the system parameters and the existence of external disturbances,a H_? filter based on the adaptive event-triggered mechanism is designed.Then,the exponential stability of the system is proved by using similar methods;Lastly,the simulation shows the rationality and feasibility of the filter.