文献类型：期刊文献

中文题名：基于Vague集相似度量的汽轮机故障诊断的研究

英文题名：Research on Fault Diagnosis of Turbine Based on Similarity Measures Between Vague Sets

作者：叶军[1]

机构：[1]绍兴文理学院机电系

年份：2006

卷号：26

期号：1

起止页码：16

中文期刊名：中国电机工程学报

外文期刊名：Proceedings of the CSEE

收录：CSTPCD、、EI(收录号：2006159819368)、北大核心2004、Scopus(收录号：2-s2.0-33645649778)、CSCD2011_2012、北大核心、CSCD

基金：浙江省自然科学基金项目(M603070);浙江省教育厅项目(20031118)

语种：中文

中文关键词：热能动力工程;Vague集;相似度量;汽轮机;振动故障;故障诊断

外文关键词：Thermal power engineering; Vague sets＇,Similarity measures; Turbine; Vibrating fault; Fault diagnosis

中文摘要：设备诊断技术有重要的实用价值,已成为一个研究的热点,并且以深厚的理论为基础。系统论、信息论、控制论、非线性科学等最新技术在其中都有广泛的应用。该文针对模糊集的推广形式——Vague集,描述了Vague集的含义与相似度量,并提出一种Vague集的相似度量在汽轮机故障诊断中的新方法。在汽轮机振动故障诊断中,Vague集之间的相似度量是评价待诊断检测样本接近系统故障知识的度量,某一相似度量值越大,待检测样本接近某一故障知识越好,从而根据相似度量值确定振动故障的类型。通过实例阐明Vague集之间的相似度量在故障诊断中的有效性和合理性。这种方法能给予多故障诊断的理论依据,比神经网络故障诊断方法更加合理、易用。

外文摘要：The technique of equipment diagnosis has importantly applied value and become a research hotspot. It is the base on profound theory. The latest techniques on system theory, information theory, control theory, and nonlinear science etc. all have the extensive application in fault diagnoses The meaning and similarity measures of vague sets were described as a generation of Zadeh＇s fuzzy sets vague sets. A new method of the fault diagnosis of turbine was presented in the basis of the similarity measures among vague sets. The similarity degree between the detecting sample and the knowledge of system fault is evaluated in the fault diagnosis of turbine vibration by means of the similarity measures among vague sets. Thebigger the value of similarity measure is, the better is the similarity degree between the detecting sample and a type of fault knowledge. Then, the type of vibration fault is determined according to the value of the similarity measure. Examples explain the validity and reasonability in the fault diagnosis problems based on the similarity measures among vague sets. This method can give you the theoretical foundation of multi-fault diagnosis, and is more reasonable and easier than the fault diagnosis method of neural networks.