Research On State Estimation Of Networked System Under Non-linear Interference

Networked systems are the product of the rapid development and interaction of computers,communications,and control.Compared with traditional control scheme,networked systems have a more complex model structure,more extensive spatial distribution,and better performance.The networked system is a real-time control system,which is mainly divided into four parts: the controlled object,sensor,controller and actuator,and its information transmission establishes a connection through the network.The networked system has the advantages of low cost,simple installation and maintenance,and high reliability,which are important factors for its wide application in many fields.However,the introduction of the communication network makes the system have problems such as limited bandwidth,network delay,data packet loss,and drive mode selection in the process of information exchange.State estimation is an important direction in the field of signal processing and control engineering.With the development of networked systems,state estimation methods need to be further explored and studied.Combined with actual engineering,it has high research value for the study of state estimation of nonlinear networked systems.This paper considers the influence of colored measurement noise,nonlinear disturbance,communication limitation,gain disturbance and node information coupling on the system,completes the design of the state estimator by solving the corresponding problems,and compares and verifies the results with numerical examples.The main work content is as follows:(1)Considering the impact of colored noise,nonlinear disturbance,communication limitation,and disturbance gain on system performance,the colored noise is processed by the difference between the measured values at adjacent moments,and the round-robin protocol is constructed to solve the bandwidth limitation problem in data transmission.The linear term is processed for state expansion and disturbance gain is added to ensure the non-fragility of the estimator,and the non-fragile extended state estimator is designed.By actively estimating the nonlinear term and estimating the error iteration equation,using the bounds of the dynamic change rate of the nonlinear function instead of the bounds of the function itself,the upper bound of the error covariance is obtained,real-time optimization of gain is realized,and the state estimator design is completed.Finally,the comparison of four simulation schemes verifies the non-fragility and feasibility of the estimator algorithm.(2)The robust extended state estimation problem of a nonlinear random complex network is studied.The state coupling between the system nodes is fully considered,and then the quantizer is used to process the state information of the nodes to reduce the bandwidth constraints of the system.Influence,by analyzing the uncertain terms caused by the quantization error,a nonlinear robust extended state estimator is designed,which can obtain the upper bound of the estimation error covariance while ensuring the estimation accuracy.Finally,the rationality and accuracy of the estimator are verified through simulation,and the control variable method is used to complete the proof of the relationship between the quantizer density and quantizer level of the logarithmic quantizer and the system performance.Finally,the designed algorithms are verified by MATLAB numerical examples,mainly through the control variable method to ensure that the estimator designed in this article is in the same environment with other estimators and the results are compared,and then verified by error analysis and data comparison.The algorithm in this paper has good non-fragility,robustness and accuracy.In the end,it is found that compared with other algorithms,the influencing factors considered in this paper are more comprehensive,with better effects,and more practical engineering.