Spacecraft Fault Detection and Isolation System Design Using Decentralized Analytical Redundancy

Abstract

Fault detection and isolation (FDI) functionality constitutes a critical element of spacecraft fault protection system capabilities. The FDI schemes currently implemented on board operational spacecraft suffer from a lack of systematic design methods. This leads to issues of decreased robustness. While model based diagnosis techniques can resolve a number of these issues, their operational applicability to spacecraft has been limited, largely due to an unfavourable net value proposition. This paper presents an approach integrating analytical redundancy based diagnosis into a conventional spacecraft FPS architecture. The approach centers on a novel decentralized diagnosis architecture based on analytical redundancy relations. A systematic procedure for designing such decentralized model based diagnosers for spacecraft is discussed, with a focus on the attitude and orbit control system. Analytical redundancy relation based error monitors and activation rules relying on the corresponding fault signatures are derived during the design phase. A comparison with the diagnosis functionality as currently implemented in the Cassini attitude and articulation control system fault protection is presented in terms of the design & development effort. It is demonstrated that the presented diagnoser design approach addresses several issues with the conventional methods, while having reasonable additional costs.