The reliability and avalibality of space-borne electric systems can be greatly improved by introducing advanced Prognostics and Health Management (PHM) technology.Based on synthetical analysis of current PHM techniques used in eletric systems

this paper put forward a hierarchical architecture

in which the model-based method and the data-driven method is combined

and feature extracting algorithms and automatic reasoning algorithms are fused at the top level to get the optimum health estimation.