Study On Adaptive Fault Tolerant Control Against Actuator Faults

Fault-tolerant technology is technology of the system fault tolerance, which makes the systeme can automatically diagnose and detect, and can automatically take the appropriate measures to ensure compensation for failures in the fault system, maintained to the required performance or acceptable range of technologies in case of failur. Since the implementation of long-term continuous control tasks, actuators are the most prone to failure. For fault-tolerant control system design, fault-tolerance strategy of the actuator should be considered a top priority.In this thesis, considering actuator faults including out-age, loss of effectiveness, and stuck, we design the control algorithm based on single-input single-output and multi-input single-output reference model. This control algorithm utilizes state feedback to control the model reference adaptive system. Considering the external disturbance of the system and the disturbances of actuators, which affect the performance of the system, adaptive compensations to control the system based on state feedback are designed. When the actuator fails, the contoller designed can identify the rate of the loss of effectiveness, design the parameters of the state feedback matrix through adaptive rate, and adjust compensate of the outage and the stuck of actors online. The design of this controller can ensure that the system can reach steady state after failures. Finally, the thesis gives examples to prove the effectiveness of the designed algorithm.This thesis also designs a controller for the single input single output linear system based on error feedback. The adaptive rates by analyzing the generalized error system are designed. When the actuators fail, the contoller designed identify the rate of the loss of effectiveness, adjust the estimations of the rate of the loss of effectiveness through adaptive rate, and adjust compensate of the outage and the stuck of actors online. The design of this controller can ensure that the system can reach steady state after failures. Finally, examples to prove the effectiveness of the designed algorithm are gived.