小卫星电源系统的分层模糊符号有向图法研究  

Fault diagnosis of small satellite power system based on hierarchical fuzzy signed directed graph

作　　者：付宇 崔玉福[1] 李志刚[1] 刘元默[1] FU Yu;CUI Yu-fu;LI Zhi-gang;LIU Yuan-mo(DFH Satellite Co.LTD.,Beijing 100094,China)

机构地区：[1]航天东方红卫星有限公司

出　　处：《电源技术》2020年第2期243-248,共6页Chinese Journal of Power Sources

摘　　要：符号有向图(SDG)模型的故障诊断方法被应用于小卫星故障诊断,但由于搜索空间大、诊断精度低等缺陷,急需改进现有模型。以小卫星电源系统为应用背景,搭建系统的SDG模型;引入分层模型,缩小搜索空间;由软件建立电源系统仿真模型,植入单体蓄电池短路故障得到正常、故障数据,计算实时节点故障隶属度,完成该故障时刻的分层模糊SDG模型的建立。通过基于模型的诊断方法推理出故障源和异常节点故障严重程度,诊断结果与假设相符,初步证明模型和算法具有正确性。通过对系统分层和引入节点隶属度,解决原模型存在的只能给出故障源而无法得到故障可能性大小的问题,实现对系统的定量解析与动态描述。Signed Directed Graph methods are mostly applied to on-orbit fault diagnosis of small satellites. However,due to the defects of large search space and low diagnosis accuracy, it’s urgent to improve the model. Taking the small satellite power system as the application background, Signed Directed Graph model was built. Then the hierarchical model was introduced to narrow down the search. Next, the simulation models needed to be established through SIMULINK, while assuming one of the battery unit short-circuited to obtain instantaneous fault data, thus membership degree of nodes were calculated by the ideal and fault data. At last, the fault was diagnosed to prove the correctness of the improved model. The purpose was to solve the existing problem that only alternative fault sources could be given but its possibilities of the failure could not be given through introducing hierarchical algorithm and membership degree, and to achieve more accurate and dynamic description of system characteristics.

关 键 词：故障诊断方法 小卫星电源系统 分层SDG模型 实时隶属度 SIMULINK仿真模型 

分 类 号：TM91[电气工程—电力电子与电力传动]