Distributed State Estimation Of Nonlinear Systems Under Random Incomplete Information

The rapid development of microelectronics,communication and sensor technology promotes the emergence of a new information acquisition and processing platform,sensor network.Sensor network has a wide application prospect in military investigation,space detection,medical care,environmental monitoring,traffic control and other fields.It has been regarded as the second largest network after the Internet in the world,and is listed as one of the four global high-tech industries in the 21st century.As a new information acquisition and processing platform,sensor network has the advantages of diverse information acquisition,strong fault tolerance and high monitoring accuracy.Sensor networks,however,are usually work in bad or dangerous environment,the sensor nodes to collect information vulnerable to all kinds of incomplete information such as time delay,missing measurement,fading measurement and the influence of network attack,and the occurrence of these incomplete information is usually arbitrary and random,it certainly will influence the accuracy of measurement nodes,and even affect the entire network measurement results.Therefore,how to design distributed state estimation algorithm with high estimation accuracy and good real-time performance is an important research direction of sensor networks.Based on control theory,matrix theory and other professional knowledge,this paper studies the distributed state estimation problem of nonlinear systems under the influence of random incomplete information.The main work is divided into the following three parts.(1)The distributed resilient state estimation problem for a class of nonlinear systems with randomly occurring communication delays and missing measurements in sensor networks is concerned with.In this part,a new sensor model is proposed to describe the communication delay and missing measurement in a unified framework,and two Bernoulli distribution sequences are introduced to express the random occurrence of the above phenomena.At the same time,the estimator gain is allowed to fluctuate within a certain range.Meanwhile,the estimator gain is allowed to fluctuate within a certain range.Based on the developed model,a novel Lyapunov–Krasovskii functional with multiple delay information terms is constructed,then the stochastic analysis technique and the extended integral inequality are used to deal with the functional derivative.Consequently,the existence conditions for the required distributed estimator are established to ensure that the estimation error system is asymptotically mean-square stable and satisfies the prescribed H_∞performance constraint,and the desired gain of distributed resilient estimator is also solved by linearizing the nonlinear terms.(2)The distributed resilient state estimation problem of nonlinear discrete systems in sensor networks is investigated.In this part,the system model under consideration involves three phenomena of incomplete information:randomly occurring nonlinearities,fading measurements,and random gain variations.The probabilistic characteristics of the above phenomena are depicted by three sets of independent random variables which are more general than Bernoulli random variables.Based on the above model,by applying Lyapunov functional approach and random distribution solution method,the asymptotic stability in the mean square sense of the estimation error system with a given H_∞attenuation level is proved.Further,the estimator parameters are solved by introducing a novel linearization method.(3)The distributed state estimation for mixed delays system under unknown attacks is dealt with.This part innovatively builds a new multi-channel random attack model,where network attacks are considered to exist in three channels:the target-to-sensor channel,the senor-to-sensor channel and the sensor-to-estimator channel.In above model,transmitted packets are allowed to be attacked multiple times simultaneously,and when they are successfully attacked,the transmitted information is modified.Besides,the topology of the sensor network is considered to change dynamically according to the Markov chain.Based on the newly established distributed estimation model,the estimation error system is proved to be asymptotically mean-square stable under a given H_∞anti-disturbance index by using Lyapunov Theory and stochastic analysis technique,finally,the estimator parameter matrices are solved utilizing linearization method.In this paper,there are 19 figures,4 tables,and 141 references.