Abstract
In this paper, we propose a new fault estimation observer for nonlinear Markovian jump systems. Following the interval type-2 fuzzy logic rules, the original system can be approximated as an interval type-2 fuzzy Markovian jump system. For such a system, we develop a new intermediate fault estimation observer, in which a variable is introduced to estimate actuator faults. Significantly, a larger design freedom can be achieved by providing the observer gain matrix in the new form. To further evaluate the proposed method, multiple fault cases are considered. The proposed method proves effective in the simultaneous fault estimation for both actuator fault and sensor fault. Furthermore, two examples are simulated to numerically verify the feasibility of the given method.