Several Kinds Distributed Variance Constraint Filtering For Discrete Stochastic Systems Over Sensor Networks

With the rapid development of computer technology,digital signal processing technology and microelectronic technology,sensor network has been developed rapidly.Sensor network has been widely used in the information transmission of control system,and brings new problems to the research of control theory,such as missing measurement,sensor delay and so on.In this paper,the distributed variance constraint filtering problem of discrete stochastic systems over sensor networks is studied by means of variance constraint method.The main contents are summarized as follows:Firstly,aiming at the distributed constraint filtering problem of time-varying discrete stochastic system with missing measurement,a new distributed filter is designed in which the available measurement information of each sensor node communicates with its neighbors,considering the interference of randomly occurring nonlinearity,multiplicative noise and additive white noise.By solving two recursively Riccati-like difference equations,the upper bound of the covariance of the filter error was obtained,and the gain parameters of the filter were designed by minimizing the upper bound.A sufficient condition for asymptotically bounded upper bound of filter error covariance is established by mathematical induction.The effectiveness of the proposed distributed filtering algorithm is verified by numerical simulation.Secondly,resilient distributed filtering problem is studied for a type of discrete-time stochastic systems subject to successive packet dropouts.The phenomenon of the successive packet dropouts for each sensor is modeled via a sequence of uncorrelated random variables obeying the Bernoulli binary distribution law.Attention is focused on the design of a resilient distributed filter and an upper bound for the filter error covariance is presented.The desired filter parameters can be obtained by a novel sparse matrix simplification method.Furthermore,a sufficient condition is established to guarantee the asymptotic boundedness of the upper bound of the filtering error covariance.Finally,a simulation is provided to illustrate the usefulness of the proposed filter approach.Thirdly,robust resilient distributed filtering problem is studied for a type of time-varying stochastic systems subject to randomly occurring uncertainties and probabilistic sensor delays.The phenomenon of the probabilistic sensor delays for each sensor is modeled via a sequence of uncorrelated random variables obeying the Bernoulli binary distribution law.Attention is focused on the design of a robust resilient distributed filter and an upper bound for the filter error covariance is presented.The desired filter parameters can be obtained by a novel sparse matrix simplification method.Furthermore,a sufficient condition is established to guarantee the asymptotic boundedness of the upper bound of the filtering error covariance.Finally,a simulation is provided to illustrate the usefulness of the proposed filter approach.