Research On Data-Driven Fault-Tolerant Control Method Based On Adaptive Dynamic Programming

With the sustainable development of science and technology,the scale and complex-ity of the industrial systems are also increasing.It is increasingly difficult to obtain the accurate mathematical model of the system by using the mechanism modeling method,and it usually takes considerable time and effort.On the other hand,one of the most important features for many industrial systems is the large amount of the offline/online input/output data generated by the system itself,which inspires people to think about how to use the real system data(instead of the mechanism model)to design the corresponding data-driven control schemes.For linear systems and affline nonlinear systems,with the help of data-driven policy iteration algorithm and a considerable amount of real system data,adaptive dynamic programming(ADP)approaches can be used to approximately get the optimal control policy,without requiring the a priori knowledge of the system dynam-ics.Nevertheless,the data-driven ADP approaches have not been applied to fault-tolerant control(FTC)systems.Based on the previous results,this dissertation is concerned with the data-driven FTC problem for linear continuous-time systems,linear continuous-time state-constrained sys-tems,linear multi-agent systems and affline nonlinear systems by using ADP.The AD-P techniques are combined with robust adaptive control and adaptive FTC approaches,which lead to a novel data-driven FTC method.The theoretical proofs are given for the main results,and the simulation experiments are carried out for the rocket fairing structural-acoustic model,the B747-100/200 aircraft model,the vertical taking-off and landing aircraft model,and so on.The simulation results show the effectiveness of the proposed approaches.The full text of this dissertation is divided into eight chapters,and the main contents of each chapter are given as follows:Chapters 1-2 systematically introduce and analyze the background and development of the FTC and its related control methods.The preliminary knowledge and research methods related to this dissertation are also provided.Chapter 3 investigates the FTC problem for continuous-time linear systems with pa-rameter uncertainties,external disturbances and actuator faults.A new data-based FTC scheme,which does not rely on the system matrices,is proposed in a parameter-dependent form.The time-varying parameters are adjusted online based on an adaptive method to compensate automatically the uncertainties,disturbances and actuator faults.The time-invariant parameter solved offline by using real system data is introduced for helping to stabilize the system.Furthermore,it is proved that all signals in the resulting closed-loop system are uniformly bounded and the states converge asymptotically to zero.Com-pared with the existing results,the proposed approach is data-based and it is easier to implement.Finally,simulation results on a rocket fairing structural-acoustic model and a B747-100/200 aircraft model are provided to show the effectiveness of the proposed approach.Chapter 4 studies the robust adaptive FTC problem for state-constrained continuous-time linear systems with parameter uncertainties,external disturbances and actuator fault-s.By incorporating a barrier-function like term into the Lyapunov function design,a novel data-driven FTC scheme is proposed in a parameter-dependent form and the state constraint requirements are guaranteed.The time-varying parameters are adjusted on-line based on an adaptive method to prevent the states from violating the constraints,and compensate automatically the uncertainties,disturbances and actuator faults.The time-invariant parameters solved offline by using data-based policy iteration algorithm are introduced for helping to stabilize the system.Furthermore,it is proved that the states converge asymptotically to zero without transgression of the constraints and all signals in the resulting closed-loop system are uniformly ultimately bounded.Finally,the sim-ulation experiments are carried out on a rocket fairing structural-acoustic model and a B747-100/200 aircraft model,and the simulation results demonstrate the effectiveness of the proposed scheme.Chapter 5 is concerned with the guaranteed cost consensus problem for linear multi-agent systems with actuator faults,where the system matrices are unknown.For the lead-erless case with actuator faults,a global performance index is constructed by all states and control inputs of all agents,where consensus regulation performances and control energy costs are considered simultaneously.Then,based on the relative state information of neighboring agents,a distributed cooperative guaranteed cost controller is designed,which not only makes the consensus problem solvable but also provides an upper bound of the given global performance index even in the presence of actuator faults.Extensions to the leader-following case with actuator faults are further studied.It is worth mention-ing that the parameters in the proposed controller are determined by using real system data.Finally,simulation results on a rocket fairing structural-acoustic model and a ver-tical taking-off and landing aircraft model are provided to show the effectiveness of the proposed method.In Chapter 6,the guaranteed cost FTC problem for unknown nonlinear systems with loss of actuator effectiveness faults is investigated using the ADP algorithm.Initially,by modifying the cost function to account for actuator faults,the problem is transformed in-to an optimal control problem of the nominal system.Subsequently,by using an existing data-driven policy iteration algorithm to solve the corresponding optimal control problem,a guaranteed cost controller is constructed approximately.Furthermore,a rigorous proof is given to show the convergence of the aforementioned policy iteration algorithm while taking the neural network approximation errors into consideration.Finally,two simula-tion examples are provided to demonstrate the effectiveness of the proposed approach.Chapter 7 investigates the robust adaptive FTC problem for unknown affine nonlinear systems with actuator faults including stuck,outage,bias and loss of effectiveness.A new data-based FTC scheme is proposed.It consists of the online estimations of the upper bounds and a state-dependent function.The estimations are adjusted online based on an adaptive method to compensate automatically the actuator faults.The state-dependent function solved offline by using real system data is introduced for helping to stabilize the system.Furthermore,it is proved that all signals in the resulting closed-loop system are uniformly bounded and the states converge asymptotically to zero.Compared with the existing results,the proposed approach is data-based.Finally,a simulation example is provided to show the effectiveness of the proposed scheme.Chapter 8 summarizes the results of the dissertation and points out the future research topics.