State Estimation And Synchronization Control Of Networked Systems

In recent years,networked systems have attracted much attention because of their superiority in robustness,flexibility and reliability.Networked system is composed of some distributed network nodes.Data transmission and resource sharing can usually be carried out between the connected nodes.The whole system can complete some tasks,such as signal filtering,cooperative operation and optimal scheduling.Networked systems have been widely used in advanced control of complex industrial processes,cooperative formation of unmanned vehicles or unmanned aerial vehicles,smart grid,intelligent transportation and smart home.Therefore,this dissertation focuses on the state estimation and synchronization control of networked systems,concerning wired transmission and wireless communication,discrete-time systems and continuous-time systems,deterministic systems and stochastic systems.The specific research contents are summarized as follows:· Firstly,state estimation problem of networked systems under communication constraints and data quantization is studied.A class of linear time-varying stocha-stic systems whose observations are affected by superposition states is considered.At each time step,only one transmission node is allowed to transmit data,and all transmission nodes are scheduled according to a Markov protocol.Uniform quan-tizer is used for data quantization.A recursive Kalman filter is designed by using orthogonal projection rule and innovation analysis method.The validity of the proposed filtering algorithm is demonstrated by simulation examples.· Secondly,the state estimation problems of networked systems with multiple pack-et dropouts and communication constraints are studied.The system studied is a kind of linear time-varying stochastic systems disturbed by superposed noises.We modeled multiple packet dropouts as Bernoulli stochastic processes,and ad-dressed them in parallel with Markov processes that existed at the same time.Then we designed recursive optimal filter estimators and smoothing estimators using innovation analysis method and orthogonal projection rule.The superiority of the proposed state estimation algorithm is verified by numerical simulation.· Then,the problem of state estimation for networked systems with multiple de-lays and multiple packet dropouts is studied.Constant measurement deviations are modeled into time-varying state equation.The orthogonal projection rule and the reorganized innovation analysis method are used to solve the problems,and we design a Kalman-like filter by using multiple Riccati equations of the same di-mensions.The validity of the designed filtering algorithm is proved by simulation examples.· Next,the synchronization control of networked systems with communication de-lay is studied.By using the aperiodic sampling control technology,we focus on the synchronization control of continuous-time networked systems.The sampled-data networked systems can be reconstructed into continuous systems by using the input delay method.The introduction of Bessel-Legendre integral inequality can further reduce the conservativeness.Then two sufficient conditions for the syn-chronization of networked systems are derived,and the corresponding controller design scheme are given.The effectiveness of the designed synchronization con-trol algorithm is verified by numerical simulation examples.· Finally,the synchronization problem of unmanned ground vehicles is studied.By combining zero-order holder,we adopt a more flexible non-periodic sam-pling data synchronization control law to study networked systems with nonlin-ear dynamics.In order to design a sampling controller with constant time delay,a new Lyapunov function is proposed based on the extended Wirtinger inequality.We establish sufficient conditions for the synchronization of networked systems.Based on the networked unmanned ground vehicle systems,the experimental re-sults in the actual scene are given,which proves the practicability of the synchro-nization scheme.