基于模糊综合评判的试验台故障诊断方法

Fault diagnosis method of ground-testing bed based on fuzzy comprehensive evaluation

提出了基于数据融合的模糊综合评判试验台故障诊断方法,该方法利用模糊输入向量和模糊关系矩阵的广义合成运算和最大隶属度原则,根据各故障原因与故障征兆之间不同程度的因果关系,在综合考虑所有征兆的基础上,对试验台进行故障诊断,并且利用模糊综合评判系数,给出故障诊断结果的可信度等级。通过对某型试验台氧系统14种故障的诊断结果验证了该方法有效性,A级正确率高于90%。

This paper proposes a fault diagnosis method of ground-testing bed based on fuzzy comprehensive evaluation and data fusion. According to the different levels of causal relationship between the cause of the malfunction and fault symptoms, this method uses the fuzzy input vector and the generalized fuzzy relation matrix synthesis operation and maximum membership degree principle to implement fault diagnosis on the basis of comprehensive consideration of all the signs of the ground-testing bed. And the credibility level of the fault diagnosis results is given by using fuzzy comprehensive evaluation factor. Simulation diagnosis re- sult of a ground-testing bed oxygen system with 14 kinds of fault modes demonstrates the effectiveness of this method, with A-level accuracy more than 90%.