Adaptive Fault Tolerant Control Design For Uncertain Systems

As modern control systems become more and more sophisticated and complicated,reliability and safety of the systems has received increasing attention.However,when the actuators,sensors or other parts of the system fail,the closed-loop system often has no expected performance and even be unstable.Fault tolerant controller design is one of the effective methods to solve this problem,which can make the closed-loop system stable with an acceptable performance in both normal and fault cases.Meanwhile,the existences of parameter uncertainties and external disturbances greatly increase the complexity of system models,and the robust control problem is necessary to be fully taken into account in the theoretical study.On the other hand,Cyber-Physical Systems(CPS)are becoming the core technology system to support and lead the new round of industrial change in the world.Since the CPS have the weakness that they are vulnerable to the network attacks,the robust fault tolerant control problem for CPS is very important and valuable.In this thesis,the adaptive fault-tolerant control problem for uncertain linear systems against time-varying actuator faults has been dealt with by fully using system information.Furthermore,a novel fault-tolerant controller for CPS is designed when sensors and actuators suffer network attacks simultaneously.Compared with the results in previous literature,the requirements of the attack information have been reduced.In view of the fact that sensor attacks may cause state information to be contrary,a corresponding fault-tolerant control strategy is proposed.The main contents of this thesis are as follows:In Chapter 1-2,the current situation and research methods in fault-tolerant control area and CPS area are summarized.Moreover,some preliminaries are also given.In Chapter 3,an adaptive fault-tolerant controller design for uncertain systems with time-varying actuator faults is presented.The actuator faults considered are more general,including time varying loss of effectiveness,outage,and stuck.By introducing an exponential decay function into the Lyapunov function,a fault-tolerant controller is designed by using adaptive technique with projection.Since the restrictions on lower bounds of some specific fault parameters are reduced,the allowable fault types are extended.Some simulation results about a rocket fairing structure acoustic model are carried out,and the effectiveness of the proposed method is further verified.In Chapter 4,the fault-tolerant control problem of CPS under sensor and actuator network attacks is investigated.By constructing new Lyapunov function and combining fuzzy logic approximation technology with adaptive method,the upper bounds on unknown parameters are estimated online,which relaxes the requirement of actuator attack information.For the linear CPS with uncertainty,external disturbance and actuator fault,a security control stagey has been designed.On the basis of Chapter 4,we further consider the fault-tolerant control problem of CPS under the condition that the sensor attack may make state information lead to the opposite direction.The Nussbaum function method,which is very effective in dealing with the unknown control direction,has been incorporated into the design process of adaptive fault-tolerant controller.The asymptotic stability of the closed-loop system under network attacks,actuator faults,uncertainty and perturbation is guaranteed.The simulation results show the feasibility and effectiveness of the proposed method.Finally,the main contributions of this thesis are summarized and some prospects for further research are also given.