Study On Adaptive Fault Tolerant Control For Uncertain Systems

Based on the adaptive control theory and related techniques, this paper addresses the problem of fault tolerant control(FTC) for uncertain systems with actuator faults. Adaptive control, control allocation, integral sliding mode control(ISMC), Lyapunov direct methods, etc., are utilized according to the features of uncertainties as well as system types for the design of adaptive FTC scheme, which is expected to be capable of dealing with failures and system uncertainties simultaneously.Firstly, an adaptive compensation control scheme for a class of discrete-time linear systems with partial loss of actuator effectiveness and parametric perturbations is proposed. A unifying framework for adaptive model parameter identification and fault estimation is designed with the construction of parameter estimator, which enables a simple and rapid adjustment. With these estimates fed back to the system, an adaptive controller is constructed to achieve a desired tracking performance.On this basis a control allocation based FTC scheme is proposed for over-actuated discrete-time systems. The baseline control law and allocation controller are designed separately. The former is designed to achieve a desired tracking performance, using indirect adaptive methods, while the latter is developed directly. More specifically, the difference between the input and output of actuators is used to adjust the control allocation matrix online. Then an adaptive control allocation is realized for the compensation of uncertainties, actuator faults and saturations.Secondly the adaptive FTC problem for linear parameter-varying(LPV) systems is discussed. Adaptive parameter updating algorithms based indirect adaptive control law is given after taking corresponding algebraic transformations and project algorithms, which are adopted for the combination or classification of uncertainties. Thus the controller can be adjusted on-line to ensure the system a satisfying tracking performance with failures and uncertainties compensated simultaneously. The sufficient condition of stability is derived which meets a Riccati inequality instead of a parameter-dependent Riccati inequality, which simplified the process of design and calculation.Finally, ISMC based FTC for systems with non-structuring uncertainties is studied. With estimates of fault parameters added in the design of conventional integral sliding manifold, an adaptive controller is constructed thus the control system has an automatic failure compensation capability. As for un-modeled dynamics, disturbance, etc., the robustness is enhanced through the combination of integral sliding mode control and other robust techniques.