This paper extends and improves the existing result on fault-tolerant control (FTC) of dynamic systems with actuator failures and external disturbances in several aspects. More specifically,the proposed method does no...This paper extends and improves the existing result on fault-tolerant control (FTC) of dynamic systems with actuator failures and external disturbances in several aspects. More specifically,the proposed method does not involve solving Lyapunov equation that contains time-varying and unknown variables associated with actuator failures; one does not need to analytically estimate the bound on the actuator failure factors in designing and implementing the proposed control scheme; the developed FTC is able to attenuate both bounded and unbounded external disturbances under actuator failures. To some extent,the results presented here include the existing results as a special case and the resultant control algorithms are fault-independent in that there is no need for explicit fault information in terms of its magnitude (size),or time instance of the fault occurrence,thus,is more user-friendly for control design and more feasible for implementation as compared with the existing work.展开更多
基金Supported by National Natural Science Foundation of China(60974052)Program for Changjiang Scholars and Innovative Research Team in University (IRT0949)Beijing Jiaotong University Research Program (RCS2008ZT002, 2009JBZ001, 2009RC008)
文摘This paper extends and improves the existing result on fault-tolerant control (FTC) of dynamic systems with actuator failures and external disturbances in several aspects. More specifically,the proposed method does not involve solving Lyapunov equation that contains time-varying and unknown variables associated with actuator failures; one does not need to analytically estimate the bound on the actuator failure factors in designing and implementing the proposed control scheme; the developed FTC is able to attenuate both bounded and unbounded external disturbances under actuator failures. To some extent,the results presented here include the existing results as a special case and the resultant control algorithms are fault-independent in that there is no need for explicit fault information in terms of its magnitude (size),or time instance of the fault occurrence,thus,is more user-friendly for control design and more feasible for implementation as compared with the existing work.