Research On Model Reconstruction Networked Control Method Of T-S Fuzzy System

Networked control systems,which use the communication network to form the feedback loops,have become research hotspots of automatic control due to the advantages of easy installation and maintenance,and high reliability.In this paper,the nonlinear networked control system is expressed as a T-S fuzzy form,and time delay is involved in the research process.Based on previous researches,less conservative stability analysis techniques are proposed.The primary research contents are as follows:For the nonlinear networked control systems in T-S fuzzy form,an H_?control method is studied,and an event-triggered mechanism is introduced to re-duce the frequency of control actions.First,considering the existence of delay,the closed-loop system is described by a T-S fuzzy model with an asynchronous membership function.To achieve better stability conditions,some slack matrices are brought into the analysis process through the deviation bounds of the mem-bership functions.Based on the improved Lyapunov Krasovskii Functional(LKF),the stability conditions expressed in linear matrix inequality(LMI)can be given.Then the relevant parameters of the event-triggered mechanism and con-trol gains are obtained,so that the closed-loop system reaches an asymptotically stable state,while the preset H_?performance index is satisfied.Finally,a simu-lation of a nonlinear mass spring system shows the effectiveness of the method.Focusing on the conservative problem brought about by the product terms of the membership functions,a model reconstruction method is proposed for the T-S fuzzy system in the network environment.First,the products of the membership functions are defined as the new premise variables,and using the upper and low-er bounds of them,the system model is transformed into a new T-S fuzzy model,so that the system matrices contain more information about the membership func-tions.Combined with the LKF method and the event-triggered mechanism,based on the reconstructed system,a series of stability criteria in the form of LMI are given,and the control gain matrices and the related parameters of the event-driven are obtained,so that the system after the model reconstruction becomes asymptotically stable,and the required of H_?performance is satisfied.Finally,a simulation of an engineering system shows that the method is useful,and the results of the comparison confirm the excellence of the method.