Communication-Protocol-Based H_? Estimation And Control For Discrete Networked Systems

The estimation and control of networked systems is one of the major research directions in the areas of signal processing and control.In this system,all the components are connected to a shared communication network in the form of network nodes.In order to avoid congestion in the transmission of data,the research assumption of reducing the amount of data transmitted into the communication network is adopted.The ordering or selection of the nodes for data transmission is realized through the introduction of a communication protocol and its scheduling.This article focuses on complex factors such as stochastic nonlinearity,measurement outliers and randomly occurring failures that exist in actual engineering.Considering the impact of the above factors on the system performance,this paper overcomes the challenges in the analysis and design process of networked systems under the communication protocol,and solves the H? estimation and control issues of several types of discrete networked systems under communication protocols.The research content provides guarantee for accurately obtaining system dynamics and boosting the smooth performance of the system,which has significant theoretical value.The main work of the article is discussed below.Firstly,for nonlinear time-varying systems with randomly occurring failures under the Round-Robin(RR)protocol,the H? fault estimator design algorithm is established.Data transmission between sensor and estimator is implemented through communication network with RR protocol to prevent the occurrence of data congestion.In the system model,slow change fault and sudden change fault in the reality are included in the system model.The stochastic nonlinearity is adopted with known statistical properties.By utilizing the completing squares approach and the stochastic analysis technique,the necessary and sufficient conditions are provided for the existence of fault estimator to ensure that the estimation error dynamics meets the prescribed H? performance constraint.Based on solving a group of coupled backward recursive Riccati difference equations,the time-varying gains of the fault estimator are calculated.Secondly,the outlier-resistant H? filtering issue is resolved for a type of nonlinear systems with measurement outliers,randomly occurring parameter uncertainties and randomly occurring nonlinearities under RR protocol.The RR protocol is utilized to schedule the communication between the sensor and the filter,and the dynamic saturation constraint is introduced when constructing the filter to weaken the negative impact of outliers on the system performance.The sufficient conditions are given to make sure that the error system dynamics meets the exponential mean-square stability and H? performance requirements under the scheduling of RR protocol.Based on solving the linear matrix inequality,the corresponding filter gain parameters are calculated.Finally,considering the influence from the stochastic communication protocol(SCP)and the measurement outlier,the H? proportional-integral-derivative(PID)control problem is addressed based on the outlier-resistant observer for discrete-time systems.The SCP is deployed to orchestrate the order of data transmission in the sensor-to-observer channel,and the random nature of the SCP is described by the Markov transition probability.This part is to design a type of H? PID controller design method based on the outlier-resistant observer,and the resistance of outlier is achieved by considering the dynamic saturation function in the observer.By establishing a reasonable Lyapunov function,the exponential mean-square stability and the H? performance of the closed-loop system are analyzed.Based on the orthogonal decomposition method of linear matrix inequality,the gain parameters of the constructed PID controller are designed.Simulation instances are given to verify and analyze the effectiveness of the proposed estimation and control methods.In general,the work in this paper is helpful to improve the research system of estimation and control for discrete networked systems in the case of communication protocol,and provides a theoretical foundation for handling other issues in the domains of signal processing and control.