The Use of Association Rules in The Analysis of the Trends of Alarms

Abstract

This article deals with acquired alarm logs analyses, the alarm logs from control systems, and explains the reasons for its collecting and implication of the analyses. Several existing platform solutions can be used to analyze the alarms from the historical trends, and solutions, which offer the filtering of events based on time units to obtain the data about the maximum and average number of logs. This paper contributes to this area by developing a graphical interface for a system that makes it possible to use the analysis by association rules and brings the possibility of detecting frequent and repeating patterns in acquired alarm data files.