H_? Filtering For Networked Systems With Data Packet Loss And Sensor Saturation

As a basic problem in the area of signal processing and control,filtering has aroused the interest of many scholars.Kalman filter has been widely used since it was proposed in the 1960s.The Riccati equation is used as the basic tool for Kalman filtering,which has been studied by many scholars due to advantages that its optimal filtering is recursive and easy to be applied in computers.However,it must to require the model parameters and noise estimation to be known accurately,which is not conducive to the practical application.H_? filter need not to know the statistical characteristics of noises exactly,and the H_? filter adopts the method of LMI,and the Matlab toolbox can be used to obtain the solution of the parameters.With the development of the times,the connection between network and human is more and more close,network control has become the focus of people's increasing attention,and networked control system arises at the historic moment.Networked control system(NCS)refers to the closed-loop control feedback system formed through the communication network,which connects various components,such as sensors,Controller and actuators through the network.Because of its advantages of low cost,high reliability and convenient data transmission,it has been widely used in various scenarios.However,due to the inevitable time delay and packet losses in the remote transmission of data over the network,it is difficult for the data to be transmitted intact.In the NCS,packet loss is one of the most common phenomena,it may reduce the system performance and make the stable region of the system narrow,more serious will make the system unstable.Currently,there are two main methods to solve packet loss.One is a deterministic research method,which mainly analyzes the corresponding network characteristics and establishes a system model to describe different packet loss situations.The second is a random method,which Bernoulli distribution or other markov process with finite state are used to describe packet loss process.Then an NCS model is established to study the relationship between packet loss probability and system performance.Sensor saturation usually exists in control systems.Due to the physical characteristics and technical limitations of the sensor itself,the sensor can only measure the signal strength within a certain range.The phenomenon of sensor random saturation usually occurs due to the fact that the sensor in the actual network measures the signal strength to be different.The nonlinear characteristic of sensor saturation not only reduces the estimation performance but also may lead to unexpected oscillation and even unstable dynamic behavior.Therefore,the problem has the practical significance that considering the influence of random saturation phenomenon existing in sensor measurement on network state estimation.Based on the situation of packet losses and sensor saturation,this paper researches the H_? filter design of networked control systems.The main research contents are as follows:Chapter 2 discusses packet losses in networked systems,redundant channel strategy is used to reduce the impact of packet loss,and the random saturation problem of sensor is also solved,the H_? full-order filter is designed in this chapter.According to the corresponding matrix inequality obtained by LMI method,the simulation analysis is carried out to obtain the optimal H_? performance index and filter parameters,which verifies the effectiveness of this paper.Chapter 3 considers the uncertainty probability of packet losses,the sensor redundancy strategy is adopted to solve the sensor saturation signal source problem,and the H_? filter for NCS with sensor saturation is designed.The optimal H_? performance index and filter parameters are obtained by using the parameter-dependent Lyapunov function for stability analysis,and the simulation results show the effectiveness of the proposed algorithm.Chapter 4 focuses on the nonlinear network control system with sensor saturation.Based on the T-S fuzzy method,the corresponding fuzzy H_? filter is researched.The effect of time delay and data packet loss is reduced by multi-packet compensation.The algorithm also adopts the sensor redundancy strategy to solve the signal source problem.The fuzzy rule dependent Lyapunov function is introduced to prove the stability of the system.By solving the LMI,the optimal H_? performance index is obtained,and the validity of the theory is verified by the analysis of simulation results.