Model-based Robust Fault Diagnosis for Satellite Control Systems Using Learning and Sliding Mode Approaches

Abstract

In this paper, our recent work on robust model-based fault diagnosis (FD) for several satellite control systems using learning and sliding mode approaches are summarized. Firstly, a variety of nonlinear mathematical models for these satellite control systems are described and analyzed for the purpose of fault diagnosis. These satellite control systems are classified into two classes of nonlinear dynamical systems. Then, several fault diagnostic observers using sliding mode and learning approaches are presented. Sliding mode with time-varying switching gains, second order sliding mode, and high order sliding mode differentiators are respectively used in the proposed diagnostic observers to deal with modeling uncertainties. Neural model-based and iterative learning algorithms-based online learning estimators are respectively used in the diagnostic observers for the purpose of isolating and estimating faults. Finally, conclusions and future work on the health monitoring and fault diagnosis for satellite control systems are provided.