基于加权粗糙集的代价敏感故障诊断方法.pdfVIP

第 27 卷 第 23 期 中 国 电 机 工 程 学 报 Vol.27 No.23 Aug. 2007 2007 年 8 月 Proceedings of the CSEE ©2007 Chin.Soc.for Elec.Eng. 0258-8013 (2007) 23-0093-07 TK38 TP277 A 470 20 文章编号： 中图分类号： ； 文献标识码： 学科分类号： ⋅ 基于加权粗糙集的代价敏感故障诊断方法 刘金福，于达仁，胡清华，王 伟 (哈尔滨工业大学能源科学与工程学院，黑龙江省 哈尔滨市 150001) Cost-sensitive Fault Diagnosis Based on Weighted Rough Sets LIU Jin-fu, YU Da-ren, HU Qing-hua, WANG Wei (Department of Energy Science and Engineering，Harbin Institute of Technology, Harbin 150001, Heilongjiang Province, China) ABSTRACT: In the realm of fault diagnosis, rough sets have 于加权粗糙集方法的代价敏感故障诊断能优先选取高代价 been a powerful tool to deal with the inconsistent information. 故障的关键征兆，并且使提取的规则集中高代价故障的规则 However, when the misdiagnosis costs of different faults are 具有更高的规则支持度和可信度，当诊断存在不一致的情况 unequal, classical rough sets can not acquire a satisfying result 下，该方法倾向于将故障诊断为高代价故障，从而降低诊断 due to the absence of a mechanism considering the apriori 代价。 knowledge. Through the introduction of subjective weights on 关键词：粗糙集；属性约简；规则提取；故障诊断；代价敏 data, this paper proposed a weighted rough set learning method 感诊断 to consider the apriori knowledge in rough sets, where an algorithm of weighted attribute reduction and an algorithm of 0 引言 weighted rule extraction were designed respectively. Based on 在工业生产过程中，不同故障对设备的危害程 weighted rough sets, the method of cost-sensitive fault 度通常是不同的，因此，在实际的故障诊断中