粗糙集理论和贝叶斯网络在汽轮机振动故障诊断中的应用  

Steam Turbine Vibration Fault Diagnosis Using Rough Sets and Bayesian Network

作　　者：董泽[1] 张楠[1] 韩璞[1] 

机构地区：[1]华北电力大学自动化系,河北保定071003

出　　处：《汽轮机技术》2008年第3期238-240,共3页Turbine Technology

摘　　要：针对实际的汽轮发电机组振动故障诊断中常存在冗余和不确定性信息这一问题,基于智能互补融合的思想,将粗糙集理论与贝叶斯网络有机结合在一起,提出了一种汽轮机振动故障诊断新方法。利用粗糙集信息表约简方法实现对故障特征的约简,以获取最小诊断规则。基于最小规则的贝叶斯网络模型可以有效降低网络构造的复杂性,同时利用贝叶斯网络来发现节点间的潜在关系,解决故障诊断过程中的不确定推理问题,从而提高诊断效率。最后,进行了故障实例分析,结果证明了该方法的有效性。According to complementary strategy, a new turbine vibration fault diagnosis method based on rough sets （RS） theory and Bayesian network （ BN ） is presented for fault diagnosis containing redundancy and uncertain information. Through reduction approach of RS information table to simplify fault symptoms. The minimal diagnostic rules can be obtained. According to the minimal rules, complexity of BN structure are largely decreased. At the same time, the potential relationship between nodes can be discovered by BN ,which can solve the uncertainty reasoning of fault diagnosis. Finally, the effectiveness of this method are validated by the result of practical fault diagnosis examples.

关 键 词：粗糙集理论 贝叶斯网络 汽轮机 故障诊断 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]