State Estimation Of Complex Networks With Markovian Packet Losses

It is difficult and unrealistic to measure the state information of all nodes in the some large-scale complex networks.However,network monitoring and fault diagnosis usually require the whole state information of each node in the network.Therefore,for the sake of achieving the all state variables and seize the behaviors of the whole network better,the state estimator needs to be constructed to estimate the whole state information through the output information of the measured nodes.The following are the main research contents and achievements of this paper.Firstly,the fact of random packet losses during the process of data transmission in the real networks will undoubtedly exert a negative impact on the state estimation of the complex networks.Different from the most existing research on complex networks with Bernoulli packet dropout,the Markov chain is used to describe the random packet losses.This chapter is concerned with the problem of state estimation for discrete-time output coupled complex networks with Markovian packet losses,which aims to reduce the influence of packet losses on state estimation.By applying the Lyapunov functional method combined with the stochastic analysis approach,a sufficient criteria for state estimation is established in terms of linear matrix inequalities(LMIs).Through a numerical simulation example,it is verified that the designed state estimator can effectively estimate the state information of the network.Secondly,for an actual state estimation system,not only can we estimate the state of the network,but also meet the practical requirement of the relevant performance indicators.Therefore,on the basis of the first conclusion,this chapter is further considered for the issue of guaranteed cost state estimation for the complex networks with Markovian packet losses.Not only the dynamics of the estimation error is asymptotically stable in the mean square,but also the guaranteed cost performance requirement is satisfied.So as to reduce the cost of the controlling system and improve the performance of the state estimator.By applying the Lyapunov functional approach and the stochastic analysis method,an effective design criteria is established for the guaranteed cost state estimation in terms of LMIs.Two simulation examples are supplied to certify the validity of the proposed scheme.Thirdly,the method of pinning control is applied on the state estimation of complex networks with random packet losses in the chapter.That is to say,controlling only a small part of nodes on the state estimator to achieve the states information of the entire network,which would reduce the network control cost and have great value of practical engineering.By using Lyapunov stability theory and stochastic analysis theory,the design criteria of the pinning state estimator is proposed.At last,two simulation examples are provided to show the number of nodes that need to be controlled in detail on the state estimator,which to achieve the global control effect for the network.It is also demonstrated that the designed state estimator based on the pinning control method is effective and applicable.