基于模糊SDG模型的航天器故障诊断方法研究

基于符号有向图(SDG)的故障诊断具有良好的完备性、易于解释性,但是其不足为分辨率差,因此,提出了基于模糊SDG模型和模糊推理相结合的半定量故障诊断方法。将节点变量变为模糊变量使节点承载了更多的定量信息,节点间的定性关系通过模糊关系矩阵来表达,通过模糊推理判断相容支路找出故障源候选集合,通过部件故障概率和传播故障权重对候选故障源进行故障可能性的排序。最后应用该方法建立了某卫星一次电源系统的诊断模型,并进行了故障诊断仿真,结果表明该方法有效地提高了诊断的分辨率,适用于航天器在轨故障诊断。

Research on Fault Diagnosis Approach for Spacecrafts Based on Fuzzy SDG Model

The fault diagnosis method based-on signed directed graph(SDG) has better completeness and detailed explanation facility,but lower resolution lowers efficiency of diagnosis,because the cause-effect relationship only by positive and negative influence of directed edge is insufficient,so the novel qualitative model based on integration of the advantage of SDG and fuzzy reasoning is given.The algorithm is presented to search the set of failure source candidates with fuzzy consistent branch by fuzzy reasoning.Furthermore,the candidates in the set is ranked according to the possibility rate defined as the product value of the failure propagation weight and failure rate of the failure source candidates.The primary power system of some satellites was modeled with the approach,the simulation diagnosis results show that the reason algorithm was validated and the diagnosis result improved diagnostic resolution.The approach is feasible to be applied for on-board diagnosis for spacecrafts.