摘要
针对水电机组常规振动故障诊断分类器不能反映分类中的不确定信息的不足,提出一种基于粗糙集的一对一(1-v-1)多类支持向量机分类方法。该方法充分利用粗糙集对不确定、不完整数据和复杂模式的良好刻画能力及支持向量机优秀的泛化能力,应用粗糙集最核心的思想:上、下近似来描述支持向量机分类结果。结合1-v-1方法实现支持向量机的多类分类,导出多类分类时样本的上、下近似和边界区域的集合表示,并以规则的形式对分类器进行描述。用所提方法对国际标准测试数据进行实验,并应用于某水电厂机组振动故障诊断。所得结果与单纯1-v-1多类支持向量机方法比较,结果表明该分类器具有规则简洁、分类阶段所需存储空间小,能够反映故障模式分类中的不确定信息等优点。
The traditional vibrant fault diagnosis classifier of hydro-turbine generating unit（HGU） can’t reflect the uncertain information in fault pattern recognition.To overcome this disadvantage,a novel classifier based on rough set（RS） and 1-v-1 multiclass support vector machine（SVM） were introduced.The proposed method takes full advantages of RS and SVM.The essential ideas of RS was used：upper approximation,lower approximation to describe the classification results of SVM.Then 1-v-1 method was used to realize the multi-class SVM classification.The set expression of upper approximation,lower approximation and boundary region in multi-class SVM classification was deduced,and the rules of the proposed classifier were extracted.At last,the method was successfully applied in analyzing an international standard data set,as well as diagnosing vibrant faults of a HGU.The results show that the proposed classifier has high classification reliability and lower requirement of memory space in operation stage,and can reflect the uncertain information of fault classification.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2010年第20期88-93,共6页
Proceedings of the CSEE
基金
"十一五"国家科技支撑计划重大项目(2008BAB29B08
2008BAB29B05)
科技部水利部公益性行业科研专项经费项目(200701008)~~
关键词
水电机组
粗糙集
支持向量机
故障诊断
hydro-turbine generating unit
rough set
support vector machine
fault diagnosis
