摘要
采用多变量决策树方法对配电网故障进行分类;运用粗糙集理论对故障特征进行提取,构造了多变量决策树;利用粗糙集具有较强的处理不确定和不完备信息的能力,对决策表的条件属性进行约简处理;同时,利用决策树具有快速学习及分类的优势对约简后的决策表进行诊断规则提取。最后用实际的配电网模型对该方法进行了验证。
With the topliss decision approach,the stoppages in distribution network was classified,then the characteristics of the stoppages by rough set theory was collected,and the multivariable decision tree was constructed.Rough sets have the great ability of dealing with indefinite and incomplete information,so reduction disposal of the condition properties in the decision tables with rough sets was done,and then the diagnosis regulation after reduction disposal can be collected easily by decision tree which is fast enough to learn the information and classify it.This approach is verified in a real distribution network model.
出处
《低压电器》
北大核心
2009年第21期44-47,共4页
Low Voltage Apparatus
关键词
配电网
故障定位
粗糙集
决策树
distribution network
fault location
roughness set
decision tree
