粗糙集理论在水轮发电机组故障诊断中的应用

Application of Rough Set Theory in Hydrogenerator Units Fault Diagnosis

为了对水轮发电机组故障诊断过程中大量的冗余信息特征进行压缩,提高诊断的效率,本文将粗糙集理论引入到水轮发电机组故障诊断中。并采用基于遗传算法的粗糙集知识约简方法,对水轮发电机组故障信息进行压缩处理。通过对具体诊断实例分析,结果表明:该方法能够在保证故障分类结果不变的情况下,有效剔除具有冗余信息,找出对故障分类起主要作用的特征,减少了对诊断信息的需求,有效地提高了水轮发电机组故障诊断的效率。

In order to improve diagnosis efficiency and compress the many redundant features in hydrogenerator units fault diagnosis, the rough set theory is introduced. The rough set feature reduction method based on the genetic algorithm is proposed and be used to deal with the fault information of hydrogenerator units. The representative samples in hydrogenerator units fault diagnosis are analyzed. The results show that when the fault classification result is almost invariable, the redundant factors can be eliminated efficiency and the main features which is more important to the fault classification can be searched through this algorithm. There search work not only reduces the requirement of diagnosis information, but also improves efficiency of the hydrogenerator units fault diagnosis.