摘要
提出了基于时间窗的序列模式挖掘算法,并将其应用于电力系统警报信息的智能处理,由此产生的规则可以补充故障诊断专家系统的知识库,为其提供警报间的时序规则知识,从而更加有效地指导电力系统的故障定位与诊断。通过一个算例说明了所提出的算法的可行性和有效性。
This paper presents a sequential pattern mining algorithm based on the time-window and applies it to mine rules of alarm in power systems. The experiments show that many interesting time-series rules can be efficiently obtained. Furthermore, this kind of knowledge can be used to guide the intelligent fault localization, diagnosis and forecasting. The feasibility and validity of the method proposed is shown by case studies.
出处
《电力系统自动化》
EI
CSCD
北大核心
2005年第13期49-52,共4页
Automation of Electric Power Systems
关键词
警报处理
序列模式
频繁模式增长
时间窗
Alarm systems
Data mining
Intelligent control
Pattern recognition
