摘要
基于决策支持技术和贝叶斯判别分析法,提出了基于特征值的电能质量分类方法,该方法使得分类结果通过电能质量实体的治理方式来表达。首先,依据电能质量监测中所涉及的众多指标进行分解,结合参考型特征样本集合和专家经验,形成了针对问题类型的规则决策治理建议库;然后,采用数学变换算法取得电能质量奇异属性对象;最后,基于贝叶斯判别分析法及其问题规则的分析,得出决策建议信息。实例计算结果表明,贝叶斯判别分类方法提高了电能质量测度的适应性,有利于对众多电能质量实体进行及时治理。
On the basis of decision support technology and Bayesian distinguished analytical method,the eigenvalue-based classification means of power quality was proposed,which makes the result of classification expressed by the power quality entity governable way.First of all,numerous indicators involved in power quality monitoring were decomposed,and regulation decision database in view of problem type was established by combining ensemble of referable feature samples and expert experiences.Then,singular property object of power quality was obtained by means of mathematical transformation algorithm.Finally,decision suggestion was proposed on the basis of Bayesian distinguished analytical method and analysis of problem regulation.The example computational results show that Bayesian distinguished analytical method improved the adaptability of power quality monitoring,which is favorable for governing numerous power quality entities timely.
出处
《太原理工大学学报》
CAS
北大核心
2010年第6期717-722,共6页
Journal of Taiyuan University of Technology
关键词
电能质量监测
贝叶斯判别分类法
数学变换算法
决策支持
power quality monitor
Bayesian distinguished analytical method
mathematical transformation algorithm
decision support
作者简介
作者简介：于烨（1979-），男，河南驻马店人，助理工程师，主要从事计算机应用研究，（Tel）13995283038
