Optimized State Estimation For Several Classes Of Time-varying Complex Networks With Communication Constraints

In this big data era,the complex network has gradually attracted the attention of scholars.In order to study the complex network in detail,the effective information of all nodes is often needed.However,due to the large number of network nodes,it is difficult to obtain the effective information of all nodes.Therefore,in order to study the complex network in detail,it is necessary to design the corresponding estimation algorithm to estimate the node state effectively.At the same time,considering the data exchange and transmission process in the communication channel,there will inevitably be some communication constraints,such as communication time-delay,missing measurements and so on.In the case of limited communication,how to design effective state estimation algorithm is of great practical significance,and it is also an urgent problem to be solved in the study of complex networks.In this thesis,new recursive state estimation algorithms are designed for several classes of time-varying stochastic complex networks under limited communication.The main contents are summarized as follows:Firstly,the state estimation problem is studied for a class of time-varying complex networks with randomly occurring nonlinearities and data packet dropouts.In order to save the network resources and improve the data transmission efficiency,the event-triggered mechanism is introduced.The norm bounded uncertainty is introduced to describe the modeling error of the system,and a stochastic variable obeying the Bernoulli distribution is introduced to describe the time-varying characteristics of the network topology.Based on the available information,the time-varying non-fragile state estimator is constructed.Then,the upper bound of estimation error covariance is found by using random analysis technique.Thirdly,the upper bound of estimation error covariance is optimized by designing the estimator gain matrix.Finally,the practicability of the proposed algorithm is verified by utilizing the MATLAB simulation.Secondly,the event-triggered resilient state estimation problem is solved for time-varying complex networks with coupling delay and sensor saturation.Two independent random variables,which obey Bernoulli distribution,are introduced to characterize the randomness of coupling delay and sensor saturation in complex networks.By introducing the event-triggering mechanism,the utilization of network resources can be guaranteed.Based on the measurable output information,a recursive non-fragile state estimator is constructed.Then,the upper bound of the estimation error covariance matrix is obtained by using the inequality processing technique.Next,an appropriate estimator gain matrix is constructed to minimize the upper bound.Finally,the feasibility and effectiveness of the design method are illustrated by MATLAB simulation.Thirdly,based on the outputs of partial nodes,the state estimation problem is addressed for a class of time-varying complex networks.By introducing the random variable obeying Bernoulli distribution,the randomly occurring nonlinearities in the system are characterized.In addition,in order to ensure the robustness of the state estimation method,a parameter matrix described by the norm bounded uncertainty is introduced.Based on the available information of partial nodes,the time-varying robust state estimator is constructed.Then,the upper bound of the estimation error covariance matrix is found by using the random analysis technique,and an estimator gain which can minimize the upper bound is designed.Finally,a simulation example testified by the MATLAB software is used to verify the effectiveness of the proposed algorithm.