Abstract
This article addresses the simultaneous fault estimation for a class of semi‐Markovian jump system with partially bounded known transition rates. A new variable is introduced with a dependence on output information, and then two robust variable‐based observers are designed to gauge the actuator fault and sensor fault concurrently. Significantly, the problem for partially bounded known transition rates is solved by virtue of mathematic techniques. In this way, the existing conditions of the designed observers are given in terms of linear matrix inequality. Furthermore, two examples are implemented to demonstrate the feasibility and effectiveness of the scheme.