Research On Nonlinear NCS Robust Fault-tolerant Control Based On The Dynamic Output Feedback In The State

With the development and maturity of network technology, more control systemrealize the exchange of information and the transmission of control signal amongsensors, sensors and controller by communication network, which makes theclosed-loop control, that is the networked control systems. They have manyadvantages, such as less attachment, low cost, easily install and maintain, and alsofault diagnosis, so they are gotten the favour by people. They have been widely usedin the field of aerospace, equipment manufacture, process control, traffic control,remote medical treatment and dangerous, special environment control. However, dueto the limit of network bandwidth and the unreliability of information transmission,which make the new problem of network delay, data dropout and so on, leading to theperformance of control system down, even instability. In addition，NCS have largescale, complexed structure, uncertainty and many factors of induced faults. NCS havemore safety and reliability with the aid of fault tolerant control, so they are paid moreattention by academics.In recent years, NCS research mainly focus on linear system, and more using thestate feedback,because of the factors of system structure, environment and economiccost, usually it’s difficult to measured the whole state information to the system, Sothe actual application of the state feedback controller is often limited. And fornonlinear NCS, fault-tolerant control study based on the dynamic output feedbackcontrol strategy has few reports in the literature. Based on this, this paper for a classof nonlinear NCS, considering the network delay, the influence of the lost package,based on T-S fuzzy model and dynamic output feedback strategy, Delay is processedby subsection forms, The conservatism of the results is reduced at the same time, faulttolerant control and the robust fault tolerant control is studied Systematicly, theconcrete content as follows:1) Research on integrality problems for nonlinear NCS with nominal andparameter uncer-tainnty.With nominal and parameter uncertainties of nonlinear NCS, using lyapunovstability theory, the LMI and liberty matrix method, based on the dynamic outputfeedback control strategy, for possible actuator fault and sensor fault, pushing cardthe time delay sufficient conditions of integrity and robust integrity, and throughsolving the LMIs, the design method of optimal controller is given. 2) Research on H_∞integrality problems for nonlinear NCS with nominal andparameter unc-ertaint.With nominal and parameter uncertainties of nonlinear NCS, considering thelimited energy external disturbance effect, using lyapunov stability theory, the LMIand liberty matrix method, based on the dynamic output feedback control strategy, forpossible actuator fault and sensor fault, pushing card the time delay sufficientconditions of H_∞integrity and robust H_∞integrity, and through solving the LMIs, thedesign method of controller is given, and the controller parameters are optimized.3) Research on guaranteed integrality problems for nonlinear NCS with nominaland parame-ter uncertainty.With nominal and parameter uncertainties of nonlinear NCS, Consider therequirements of the dynamic performance, using Lyapunov stability theory, the LMIand liberty matrix method, based on the dynamic output feedback control strategy.For possible actuator fault and sensor fault, pushing card the time delay sufficientconditions of guaranteed integrity and robust guaranteed integrity, and throughsolving the LMIs, the design method of optimal controller is given. And the controllerparameters are optimized.In the basis of the above-mentioned theory research, the correspondingsimulation examples are used to illustrate the correctness and effectiveness of thesimulation. In the paper, the delay of the section processing not only conform to theactual situation, but also can effectively reduce the conservatism of the results. In theprocess of proof of the application of free weighting matrix， which does’ttransformate the model of original system, and also does’t define a cross, so thegotten conditions have a more small conservative. So as to increase the feasibility ofthe controller design and the satisfaction of fault-tolerant control.