| description abstract | Actuators may fail during system operation, and actuator failures are often uncertain in the sense that it is not known when, how much, and how many actuators fail. One typical actuator failure is that some unknown inputs may be stuck at some unknown fixed values at some unknown time instants. Actuator failure may cause major problems in many critical control systems such as flight control systems. The key design issue is how a control system is intelligent enough to use remaining working actuators to achieve desired performance in the presence of unknown actuator failures. Adaptive control is an effective method for controlling systems with uncertainties including those in actuators. Adaptive actuator failure compensation, aimed at compensating such uncertainties with adaptive tuning of controller parameters based on system response error, is an important topic of adaptive control research with both theoretical and practical significance. Recently, there have been encouraging results on control of systems with actuator failures. The multiple models, switching and tuning designs (1-2), adaptive designs (3-5), fault diagnosis method (6), and robust fault accommodation (7) have been used for control of systems with actuator failures. In Refs. 6-11, we developed several direct adaptive control schemes for continuous-time linear time-invariant plants with actuator failures characterized by some unknown inputs stuck at some values not influenced by control action. Some key issues in adaptive control of systems with actuator failures, such as actuator failure models, controller structures, plant-model matching conditions, error models, adaptive laws, stability analysis, have been addressed for the considered cases: state feedback for state tracking (8,11), state feedback for output tracking (9), and output feedback for output tracking (10). | |
