Fault Detection for an Aerospace Launch Vehicle Using Sliding Mode Observer

Aerospace Launch Vehicles (ALV) are generally designed with high reliability to operate in complete security through fault avoidance practices. However, in spite of fault avoidance, fault occurring is inevitable. Hence there is a requirement for on-board fault detection and isolation (FDI) without significant degradation in the ALV performance. The robust observers are widely used in FDI due to reduction of the effect of disturbances in the FDI process. In this paper, the robust fault diagnosis observer is designed for an ALV subject to uncertainties. The linear sliding mode technique is used to design the observer for a linear time varying model of an ALV. The parameter estimation from the sliding mode scheme is compared with those generated by a nonlinear simulation and are found to provide good correlation. Then, a proposed linear sliding mode observer is employed to generate the residual as an indicator of predefined gyroscope faults.