Research On State Estimation And Filtering Methods For Complex Systems Based On Fuzzy Model

The ordered society depends on the smooth operation of various complex systems.Complex systems play a very important role in the fields of natural science and social science.However,there is no concise definition of complex systems.It is generally considered that complex systems have the following properties: nonlinear,complexity and high-dimension.Complex systems are ubiquitous in people's living,such as urban traffic network,integrated circuit system and gene regulatory network.Generally,a high-dimensional interconnected large-scaled system composed of finite coupled nonlinear subsystems is defined as a complex system.As a typical complex system,nonlinear interconnected large-scaled systems have become a hotspot of scholars because of their importance in practical application and theoretical research.In some practical complex systems,the existence of external disturbance and limited channel resources will directly or indirectly affect the dynamic performance of the complex systems.Moreover,due to the perturbation phenomena often occurring in the operation of devices and the transmission of signals,the established model needs to be more precise and close to the actual dynamic system.In addition,the existence of switched mechanism may directly or indirectly affect the dynamic performance of complex systems.It is also imminent to study complex systems with switched mechanism.Switched systems consist of a finite number of nonlinear subsystems with continuous or discrete-time dynamics and switched rules that control any particular subsystem moving along the trajectory of the switched system at any time.Because asynchronous may occur in the switched nonlinear systems.Based on the above factors,the analysis of complex system model,state estimation and filtering,and the study of transmission mechanism have high research significance and engineering application prospects.T-S model has been proved to be a universal approximator to approximate various smooth nonlinear functions.Using T-S fuzzy model to construct complex systems can solve the problem of nonlinearity of complex systems.The main work and innovation of this dissertation are as follows:(1)In this dissertation,the problem of non-fragile H_? filter for nonlinear interconnected large-scaled systems based on T-S fuzzy model is studied.The filter may have parameter perturbation due to the influence of external environment and the aged internal components.It is of great practical significance to design a non-fragile H_?filter for this problem.Because interconnected large-scaled systems are composed of a finite number of coupled nonlinear subsystems,centralized filtering methods for complex systems may cause serious congestion and packet loss in transmission channels because of the large scale and data,and it is also difficult to achieve in practical applications.In this dissertation,a distributed non-fragile H_? filter is designed for the discrete interconnected large-scale system based on T-S fuzzy model.According to the design criteria of the filter,the stability of the established filtering error system model and the performance of H_? are analysed.Based on the above analysis results,a non-fragile H_? filter design method is obtained.(2)This dissertation studies the design of event-triggered H_? distributed non-fragile filters for interconnected large-scaled nonlinear systems.Because many unnecessary data are sampled and transmitted in the traditional time-triggered mechanism,a H_? filter based on event-triggered mechanism is proposed for interconnected large-scaled systems based on T-S fuzzy model in this dissertation.Data in event-triggered mechanism can perform corresponding tasks only when certain conditions are met,thus the frequency of data updates and the occupation of unnecessary data on the channel are reduced.The simulation result shows that the proposed event-triggered mechanism improves the utilization of channel resources,and guarantees the stability of the filtering error system and the expected H_? performance.(3)The asynchronous H_? filtering problem for T-S fuzzy switched nonlinear systems is also studied in this dissertation.In practical engineering,the filters belonging to each subsystem often fail to run in time when the subsystem is switched,so this dissertation designs an asynchronous filter to deal with the asynchronous phenomenon.Secondly,By using the MDADT and the Lyapunov function methods,the switched filtering error system is uniformly asymptotic stability and has the H_? performance.The filter parameter is solved by linear matrix inequalities(LMIs).(4)In this dissertation,H_? filter based on integrated event-triggered is also designed for interconnected large-scaled nonlinear systems.In the actual operation of the system,when the system performs control tasks without disturbance,the use of event-triggered mechanism whose trigger threshold is independent of the system state may have limited effect on reducing the communication burden.Combining the event-triggered mechanism whose trigger threshold is independent of the system state and the event-triggered mechanism whose trigger threshold depends on the system state,this dissertation proposes an integrated event-triggered mechanism for interconnected large-scaled systems based on T-S fuzzy model,which can further reduce unnecessary data transmission compared with the traditional event-triggered mechanism.The simulation results show that the proposed integrated event-triggered mechanism can improve the utilization of channel resources,and ensure the stability of the filtering error system and the expected H_? performance.