摘要
根据属性约简过程中决策属性集相对条件属性集的条件熵的变化规律和属性,在分明矩阵中出现的频率作为启发式信息,提出了基于熵和属性频度的属性约简算法。在此基础上把粗糙集自动知识获取的理论应用在电力系统的变压器故障诊断。实例分析表明,该方法有效地减少了故障信息的冗余性,诊断效率高,结果易于理解,在电力系统其它领域可进行类似推广。
According to the changing tendency of the decision attributes given condition attributes and the frequency of attribute in discemibility matrix, then a new attribute reduction algorithm based on information entropy and attribute frequency in discernibility matrix is proposed, then the rough set theory of automation knowledge acquisition is applied in transformer fault diagnosis in power system, the example shows that the proposed method reduces the redundancy of fault diagnosis information with high efficiency, and easy to be understood, it can also be applied in other field in power system.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第2期418-420,共3页
Computer Engineering and Design
基金
湖南省教育厅基金项目(06C125)
关键词
粗糙集理论
分明矩阵
属性约简
条件熵
启发式信息
rough set theory
discernibility matrix
attribute reduction
condition entropy
heuristic information
