应用粗糙集理论研究介损在线监测数据规律

STUDY ON REGULARITY OF ON-LINE MONITORING tan δ BASED ON ROUGH SET THEORY

介损在线监测数据波动很大并与外界因素有关。如能从数据中挖掘出这些因素对介质特性综合作用的规律性 ,将为诊断提供有力依据 ,因此提出了一种基于粗糙集理论的介损在线监测数据规律性挖掘方法 ,将连续监测量离散化形成信息表 ,通过设置支持度与信任度对信息表进行预处理来保证监测记录的普遍性 ,再约简信息表得出绝缘特性变化的一般规律。最后 。

On line monitoring tan δ data are very fluctuant, mainly due to influences of many factors, such as voltage, temperature and humidity etc. If influence regularity of these factors can be found from large mount of the monitoring data, strong support can be provided for diagnosis. In this paper, based on rough set theory(RST), a regularity mining approach for on line monitoring tan δ data is proposed. At first discretizing continuous data to form the information table, then through setting support and confidence to preprocess the discretized data, and general insulation regularity of the analyzed equipment can be gotten through RST reduction. Finally, actual analytical results verify the effectiveness of the proposed approach.