基于改进D-S的汽轮机组集成故障诊断研究

Fault Diagnosis Method Based on Improved D-S Evidential Theory for Turbine Generator Unit

在分析与比较多个D-S合成规则的基础上,结合汽轮机组故障的特点,提出了一种基于改进D-S证据理论的集成故障诊断方法。该方法利用改进的D-S理论来表示和处理不确定的、模糊的信息,利用灰色理论和GRNN(广义回归神经网络)网络来处理证据理论中的基本概率分配问题,充分发挥灰色理论和GRNN的优点,提高故障诊断率。仿真结果表明,所提出的集成故障诊断方法能有效地诊断汽轮机组的故障,决策合理,可信度高,且能避免误诊现象,具有良好的应用前景。

Based on analyzing and comparing various D-S compound rules and combining the feature of turbine generator sets, an integrated fault diagnosis method was proposed. This method dealt with the uncertain and fuzzy information by improved D-S evidence theory, and solved the basic probability assignment of evidence theory by grey theory and generalized regression neural network （GRNN）, which made advantage of grey theory and GRNN and improves the rate of fault diagnosis. The simulation results show that the proposed method can effectively diagnose the vibration fault of turbine generator set, avoid misdiagnosis, and has good application prospects.