基于改进符号有向图模型的发动机引气系统多故障诊断方法

为了在符号有向图(SDG)模型中进行多故障诊断,提出了基于改进符号有向图(ISDG)模型的多故障诊断方法.ISDG模型满足了不完全信息条件下的多故障组合诊断的需求.通过交互式方法构建不完全信息条件下的诊断过程,利用最大增益费用比确定了最优的测试序列,实现了在多故障诊断过程中效率的提高和成本的降低.最后用交互式算法诊断某民用发动机引气系统多故障,ISDG模型能够诊断多故障,说明诊断多故障可以提高诊断效率;考虑组合逻辑后,最小费用比最大费用减小了7.25,增益费用比增大了32.2%,说明考虑组合逻辑可以减少32.2%的费用.

Multiple faults diagnosis method of engine bleed system based on improved signed directed graph

To diagnose multiple faults in signed directed graph (SDG) model, an improved signed directed graph (ISDG) multiple fault diagnosis model was proposed. The proposed ISDG model can meet the requirements of multiple faults combinations diagnosis in incomplete information conditions. The diagnosis process in incomplete information conditions can be established by interactive method. The optimum test sequence was determined by the maximal gain-cost ratio. This model improved the efficiency and reduced the cost in the multiple fault diagnosis process. Finally, the multiple faults of an civil engine bleed system were diagnosed with the interactive algorithm. The multiple faults can be diagnosed by ISDG model, so the diagnostic efficiency was improved by diagnosing multiple faults. After using combinational logic, the minimum cost is reduced by 7.25 than the maximum cost, and gain-cost ratio increases by 32.2%. These results explain that the cost can be reduced by 32.2% by using combinational logic.