Robust Failure Detection for Reentry Vehicle Attitude Control Systems

Abstract

This paper presents a robust failure detection methodology for the attitude control system of reusable launch vehicles (RLVs). In particular, we consider the problem of estimating the thrust from multiple jets e ring from an RLV reaction control system (RCS), as well as the related problem of distinguishing between failures in the RCS and the aerosurfaces. For accurately known vehicle and sensor models, the Kalman e lter provides the optimal estimate for the jet thrust, in the least-squares sense. During reentry, however, plant model uncertainties are a major problem for such a e lter as the vehicle’ s aerodynamics vary widely with rapidly changing Mach number, making gain scheduling impractical. Consequently, the Kalman e lter’ s performance degrades. Even if the Mach number were accurately known, rapid gain scheduling may not be desirable or even possible, because of the large data storage requirements it entails. Transient, robust H1 or game-theoretic e lters are proposed for nextgeneration RLV, and a prototype design is demonstrated for the Space Shuttle Orbiter’ s attitude determination system. Simulation results demonstrate that the robust e lters can be insensitive to plant model uncertainties over a much wider range of Mach numbers than a traditional Kalman e lter, while remaining sensitive to failures in the aerosurfaces and the RCS jets.