基于粗糙集理论的电力变压器故障诊断方法

鉴于电力变压器信息的不完备性及复杂性，基于粗糙集理论提出了一种能较好处理不完备信息的变压器故障诊断模型。基于对大量电力变压器故障征兆及故障类型的分析统计，利用粗糙集进行约简以获取诊断规则。文中详细阐述了在获得各类信息情况下如何利用该模型进行故障诊断；即使缺少某些关键信息时，该模型也能结合欧式距离、神经网络和模糊数学三种方法对约简进行综合匹配，再利用相应的约简及规则集作出故障诊断。该模型还可通过丰富训练样本、修正决策表的自我完善方法使诊断效果不断提高。实例也表明了该方法的有效性。

DIAGNOSTIC MODEL OF INSULATION FAULTS IN POWER EQUIPMENT BASED ON ROUGH SET THEORY

Due to the incompleteness and complexity of fault diagnosis for power transformer, a specific fault diagnostic model with self-improvement method based on the Rough Set theory is given in this paper. After the statistic analysis on the collected fault examples of oil-immersed transformer, the results were reduced according to rough set theory and then the diagnostic rules were gotten also. In this paper, how to use this model for fault diagnosis under various conditions was described. Especially lacking key information, the reduced decision table may be gotten by synthetically matching the information came from artificial intelligences, i.e. Euclid Distance, Artificial Neural Network and Fuzzy Mathematics, and then the diagnosis might be completed by the reduced decision table and the accordingly rule set. Meanwhile the effectiveness of this model can be enhanced by the self-improvement method, which is achieved by modifying the decision table through richening training sample. The case studies show it is effective and useful.