变压器故障诊断的粗集决策新方法  

A New Method of Rough Set Decision for Transformer Fault Diagnosing

作　　者：朱振玉[1] 董春辉 

机构地区：[1]黑龙江科技学院计算机与信息工程学院,哈尔滨150027 [2]北京握奇数据系统有限公司,北京100015

出　　处：《电测与仪表》2010年第7期21-23,29,共4页Electrical Measurement & Instrumentation

摘　　要：通过分析DGA新导则在实际应用存在的问题,将粗集理论与DGA相结合,提出变压器故障诊断的粗集决策方法。在保证离散化结果性能要求的前提下,使用改进的贪心算法进行连续属性的离散化。最后通过规则提取,得到变压器故障类型的决策规则,结果表明方法是有较高的准确率。The questions were found when the DGA new rules were applied in the practice.So a new method of rough set decision of transformer fault diagnosing is brought forward through combining rough set with DGA in the paper.Then better greedy algorithm is used in the course of discretization of continuous attributes in order to assure performance of result of discretization.Finally decision rules of transformer fault diagnosing are gained and in the result the new method has higher exact ratio.

关 键 词：变压器 粗集 故障诊断 DGA 贪心算法 

分 类 号：TM40[电气工程—电器]