Application of Uncertainty Reasoning Theory to Satellite Fault Detection and Diagnosis
Tianshe Yang,Zheng Xi,Lida Xu,Xiaoshe Dong,Yongxuan Huang
DOI: https://doi.org/10.1109/icsmc.2007.4414169
2007-01-01
Abstract:Reasoning theories are divided into certainty reasoning theories and uncertainty reasoning theories. Now, only certainty reasoning theories are used to detect and diagnose satellite faults. However, in practice, it is difficult to detect and diagnose some faults of the satellite automatically only by use of certainty reasoning theories. The reason is that detection and diagnosis of these faults require a rational reasoning and a fault-tolerant capability. Fortunately, uncertainty reasoning theories can meet these requirements. It is attracting attention of many experts in the space field all over the world that uncertainty reasoning theories are applied to detect and diagnose satellite faults. Uncertainty reasoning theories include several kinds of theories, such as inclusion degree theory, rough set theory, evidence reasoning theory, probabilistic reasoning theory, fuzzy reasoning theory, and so on. Inclusion degree theory, rough set theory and evidence reasoning theory are three advanced ones. Based on these three theories respectively, this paper introduces three new methods to detect and diagnose satellite faults. It is shown that the methods, suitable for detecting and diagnosing satellite faults, especially uncertainty faults, can remedy the defects of the current methods.