基于粗糙集与证据理论的凝汽器故障诊断研究

Study of Condenser Fault Diagnosis Based on Rough Sets and an Evidence Theory

针对凝汽器故障诊断问题,提出了一种基于粗糙集和证据理论相结合的故障诊断方法。利用粗糙集相对约简的不唯一性,对凝汽器故障征兆进行分类,形成不同的证据来源,既实现了证据理论对于同一事物要求有不同的证据来源的要求,又对故障征兆参数进行了降维处理,减小了网络的规模,有效缓解了由于输入参数过多给网络带来的收敛困难问题。该诊断方法将粗糙集、神经网络和证据理论有机地结合在一起,使三者优势互补,充分利用了凝汽器故障征兆的冗余、互补信息。实例证明,基于多故障诊断网络信息融合的诊断识别准确性和可靠性比基于单一故障诊断网络的诊断识别有较大的提高。

In the light of the problems relating to the condenser fault diagnosis,proposed was a fault diagnosis method based on a combination of rough sets with an evidence theory.By utilizing the non-uniqueness of the relative reduction of the rough sets,the signs of the condenser faults were classified and various evidence sources were formed.This not only meets the requirement that the evidence theory needs various evidence sources for a same matter but also in a dimension-reduction way treats the parameters repre...