Fault estimation for linear continuous-time systems with actuator failures

This paper addresses the problem of state and fault estimation for linear continuous-time systems with actuator failures. A new estimation technique is presented to deal with this design problem. In the proposed approaches, the original system is first augmented in a descriptor system, and a new sliding mode observer is introduced to obtain accurate estimations of both system states and actuator faults. A numerical example is presented to illustrate the effectiveness and applicability of the proposed technique.