DUCG-based Spacecraft Fault Diagnosis Methodology

In order to solve the problems of intelligent fault diagnosis of spacecraft, this paper builds a diagnostic model based on Dynamic Uncertain Causality Graph (DUCG). This model solves the issues with data-driven and model-based methods, such as poor interpretability, high data dependence, and low diagnostic correctness. The DUCG methodology relies on the expertise of the domain experts to demonstrate uncertainties between the spacecraft telemetry parameters and possible faults in a graphic way. Notably, DUCG exhibits high diagnostic accuracy and interpretability, even in the absence of existing fault data. The diagnostic model developed utilizing the DUCG contains a total of 29 faults and 110 telemetry parameters. Empirical findings indicate that the model achieves a remarkable accuracy rate of 100%. The research results in this paper are of great significance for the development and promotion of spacecraft fault diagnosis technology.