Application of Sequential Pattern Mining Algorithm in Alarm Information Processing of Power System
Yan Zhang · Dianli xitong zidonghua · 2005
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.